1.0 Introduction

Thus, a crucial issue when contemplating system architectures for space systems is thinking about incorporating molecular nanotechnology at some point. Very little work, however, has been done in this area. For system architects to consider the issue, they need to be provided with an understanding of the range of space systems and attendant capabilities that molecular nanotechnology could offer.

Once provided this information, aerospace systems engineers then face the classic systems architecting problem of whether or not, and how, to incorporate a technology which promises a dramatic increase in system performance, but which carries substantial schedule, cost, and technical risks. This is particularly difficult for molecular nanotechnology, since dramatic performance increases in supporting technology often require a re-architecting to properly exploit [Rechtin, 1991], [Raymond, 1951].

Indeed, the prospects for significant new architectures for space systems are particularly strong with molecular nanotechnology, as it offers not only higher technical performance on some functions, but as [Drexler, 1992a] points out, several novel functions for technical systems as well.

The focus of this research is to examine architectures for space systems that exploit molecular nanotechnology. It will be successful if it accomplishes the following objectives: 1) Clearly refute, support, or show as currently undecidable by analysis each of the major claims in the literature about systems architectures for space systems that exploit molecular nanotechnology, as listed in 3.1 below; 2) For each major Space System function listed in 2.2.2 below, develop a supported minimum estimate for the maximum performance on that function for some system architecture by using molecular nanotechnology; and, 3) Generate at least one novel system architecture for space operations which is realizable within our understanding of physical law.

Throughout this research proposal, "MNT" will be used interchangeably to mean "molecular nanotechnology."

The research will approach this topic in two major steps, first assessing the capabilities of previously defined system architectures, operational concepts and grand strategies, using the technical performance parameters estimated for molecular nanotechnology, and second, finding, generating and assessing system architectures, operational concepts and grand strategies specifically created to exploit the estimated capabilities of molecular nanotechnology.

"Grand strategy" here means the top-level approach to achieving primary goals, an understanding of why that approach is expected to be effective, and an understanding of why the goals are important and appropriate. For example, the grand strategy for Apollo was setting a spectacular and far off goal in the space arena in which the Soviet Union had been beating the US. The long-term goal removed the focus from the short term "firsts" that the Soviet Union could achieve since it was ahead, and gave the US enough time to surpass the Soviet's in space. This was important because the United States was engaged in a political and ideological struggle with the Soviets, and their highly visible space victories were significant propaganda elements in persuading people and countries more towards the Soviet Union.

Now is a good time for the proposed research. First, MNT has become sufficiently well defined technically [Drexler, 1992a] that it can be analyzed and used in synthesis and analysis, and MNT is also starting to receive recognition, such as the recent RAND manuscript [Nelson and Shipbaugh, 1995]. Second, world context has changed so dramatically that it is time to reassess our grand strategies for space [Friedman, 1995a], and insights on the long-term implications of MNT would support that reassessment . Finally, systems architecting has developed into an understandable mindset [Rechtin, 1991] which could support, and grow from, this research.

It is hoped that this research will also achieve some or all of the following: 1) Generate not just one, but several novel system architectures for space operations realizable within our understanding of physical law; 2) Create an understanding of the range of systems architectures for space operations made feasible with molecular nanotechnology; 3) Understand the competitive capabilities of these systems architectures; 4) Develop a range of concepts for space operations which use combinations of these systems architectures; 5) Understand the competitive capabilities of these operational concepts; 6) Create an understanding of the range of Grand Strategies for space operations, given the development of molecular nanotechnology, including the near and medium term actions implicit in these grand strategies; and, 7) Understand the relative suitability of these different Grand Strategies, given the development of molecular nanotechnology. At a minimum, it is hoped that this research will form a basis from which further research will eventually achieve these larger goals.

It is also desired that this research provide an illustrative example for systems architects and systems engineers of assessing the crucial relationship between emerging technology and unprecedented systems, since new technology which promises to provide a dramatic jump in system effectiveness, but which actually is the primary cause of risk on a new development, remains a troubling area in systems engineering. Toward that end, the actual methodology in progress will be carefully documented for the benefit of future researchers.

The question this research will answer is "what would MNT-based space systems roughly be capable of achieving?"

That question makes this a dissertation in systems architecting, because it requires a systems architect's focus and perspective to grapple with. In order to answer the question, one must answer "what will MNT-based space system architectures look like, and why?" It requires the ability to modify system architectures and create new ones as needed to satisfactorily exploit, at the system level, the benefits of a powerful, sometimes driving, component technology.

This research will illustrate how to grapple at the system level when presented with a potentially very powerful component technology.

This research will also help in better understanding more about the real value of MNT, if any, and help in beginning to recognize where MNT would lead to fundamental redesigns of space systems. Furthermore, if the operational utility of MNT to space, well and conservatively estimated, proves significant, this will stimulate further investigation, effort and investment, leading to more results, building on itself. If MNT does not really offer much increase in space operational capabilities, then the space community and space system architects will be able to safely put MNT to rest.

Figure 1. This proposed research fills the gap between systems architecting for space operations, and molecular nanotechnology.

2.0 Background

2.1 Overall Systems Architecture

• The "system" is what is built.
  
• The "model" is a description of the system to be built.
  
• The "system architecture" is the structure of the system.
  
• The "overall architecture" includes the structure not only of the system, but of its functions, the environment within which it will live, and the processes by which it will be built and operated.
  • [Rechtin, 1991]

Systems architecting is the structuring of a system so that it will satisfy a purpose. The term is often restricted to the creation and building of unprecedented complex systems [Rechtin, 1994], which is the most difficult case. In current systems architecting thought, the "system" is usually an aerospace product, communications system, computer hardware or software, or a manufacturing system [Rechtin, 1991]. When the system is a building or ship, the related disciplines of civil and naval architecting apply.

In this research proposal, the term "overall systems architecture" is used interchangeably to mean "overall architecture." The concept of an overall architecture was brought into focus by the seminal work [Rechtin, 1991], in which he pulled together and defined the modern concept of systems architecting [McKendree, 1993].

The "overall architecture" is an expansion of the concept of "system architecture." First, the system architecture is the arrangement of the components of a system to achieve its emergent function(s). The overall architecture adds the system's purpose, its functional structure, and how it fits into the environment over the course of its life cycle. As [Rechtin, 1991] observes, the system is embedded in larger systems, which in turn are embedded in larger systems, and many of these larger systems overlap, so an overall architecture is an intrinsically unbounded subject. Setting bounds to what one will consider in a particular case is a crucial part of the early structuring of the problem, which systems architects must do in their work.

The system architect is the agent of the client, responsible for translating the client's wishes into a conceptual design which the builder can build. This is a necessary role, but one which has often lacked formal recognition. Rechtin is a very strong advocate for having the system architect also be responsible for system certification [Rechtin, 1994].

The focus of much systems architecting work is creating the top-level conceptual design of the system. In the earliest stages of a project, this top-level conceptual model for most practical purposes expresses the systems architecture.

Systems architecting is purpose oriented [Rechtin, 1994]. It relies on a technology base for what is feasible. The operational concept describes how the system is to be used and what it is intended to do in the field, and grand strategy defines what this is supposed to achieve and why it is important. These three levels form a coarse-grained example of a hierarchy of purposes. While [Nadler, 1981] and [Nadler and Hibino, 1990] emphasize using a purpose hierarchy to select the appropriate level to focus one's attention, it is also true that to satisfy a purpose, all the levels must ultimately come to be in satisfactory accord.

[Goure, 1994] speaks indirectly of this hierarchy of supporting technology, systems, operational concepts, and grand strategy. Each level must support the level above it, and has as its purpose to satisfy the level above it. At the same time, each level can only ask the level below it to support what that level below can reasonably provide. The technologies support the system architecture, and they must match each other. The architected systems are used, often in combination with other systems, and that use is captured in the operational concept. They must match each other. Finally, the operational use implements a grand strategy, and is usually guided by an explicit grand strategy. These relationships are shown in figure 2 below.

Figure 2. The technology, system architecture, operational concept, and grand strategy, must all match.

During the conception and development of a new system, that system is in a future system state [Nadler, 1981]. One effective means of thinking about the future is scenarios [Schwartz, 1991]. Specific scenarios will include specific system architectures as elements, specific operational concepts as plot or subplots, and a specific grand strategy providing the overriding motivation. It may provide multiple grand strategies for different adversaries. Occasionally, a grand strategy may change in a scenario, that change being a major element in the plot.

The Journal that includes systems architecting in its scope is Systems Engineering: The Journal of the International Council on Systems Engineering. Some relevant subjects are also appropriate for IEEE Transactions on Systems, Man and Cybernetics [Rowe, 1965].

There are two different types of experts in systems architecting (although one person can be an expert of both types): 1) experts in studying systems architecture; and, 2) experts in creating systems architectures. In other words, there are experts on the subject, and experts in the practice, the same as the difference between a military historian and a general.

It is easier to identify experts in systems architecting, since often all one need do is select large, complex, unprecedented systems, and then answer "who was the architect?" Examples include G. Bennig and H. Rohrer, system architects for the STM [McKendree, 1989a], W. R. Schindler and J. Kline of Douglas Aircraft Company, system architects for the Delta Rocket [McKendree, 1989d], General Barker, General Morgan and Major General J. A. Sinclair, system architects for Operation Overlord [McKendree, 1989c], and Walt Disney, system architect for Disneyland [McKendree, 1989b].

Experts on systems architecting are harder to determine. Rechtin, however, clearly is one [Rechtin, 1991], [Rechtin, 1994]. Simon [Simon, 1981], Alexander [Alexander, 1964] and Klir [Klir, 1985] are also relevant experts.

2.2 Overall System Architectures for Space Operations

Section 2.2.1 lists mission areas for space operations. Section 2.2.2 lists major functions performed by space system architectures. Sections 2.2.3 through 2.2.10 discuss overall system architectures for space operations within the defined categories of mission areas. Finally the main journals in the field are summarized in 2.2.11

2.2.1 Mission Areas for Space Operations

1. Science for the sake of pure knowledge.
2. Exploration into unknown domains.
3. Exploitation of material and energy resources back to Earth.
4. Enhancement of quality of life on Earth.
5. Human colonization of space.
6. Planetary defense.
7. International competition.

Adding an additional mission area of:

0. Provide support to other mission areas

more cleanly addresses the many space functions which are repeatedly used across many mission areas.

2.2.2 Major Space System Functions

Table 1 (next page) illustrates the relationship between these major space functions and the mission areas. Each mission area has one or more primary functions which represent the realization of the mission area. For example, one set of system architectures which enhance quality of life is communications satellites. Signal relay is the function which implements their purpose. Thus signal relay is checked under Enhance Quality if Life. Most systems for space operations across the mission areas however, also require signal relay as a necessary support function, even though their major functions are something else. Thus signal relay is also checked under Support.

2.2.3 through 2.2.10 below discuss example overall system architectures within the mission areas listed across the top in table 1.

Table 1. Major Space System Functions support various space mission areas.

2.2.3 Support Architectures

Most support architectures are means of transporting payloads through space. The first such architecture is the rocket. Analyses going at least back to Goddard [Goddard, 1970] show that with a sufficiently energetic fuel, a rocket can expel reaction mass from burning, and the recoil will lift the rocket into space. Furthermore, rockets can be used to change trajectories when already in space. Unfortunately, rockets operate under the burden of carrying their own reaction mass, which must be accelerated along with the rocket and payload before being used. This gives rise to the exponential rocket equation:

(rocket mass + reaction mass used)/(rocket mass) = e^{[[Delta]]V/(Isp*g0)} (1)

"Rocket mass" includes all the accelerated mass (payload, structures, engines, unused reaction mass, etc.) except the "reaction mass used," which is the mass of material expelled to provide acceleration. [[Delta]]V is the total change in velocity. Isp is the specific impulse, a measure of the effectiveness of the engine, often interpreted to mean the number of seconds that one pound of reaction mass can deliver one pound of thrust (in the basic rocket equation above, this is assumed constant). g0 is the acceleration of gravity on earth's surface (9.8 m/s).

Example Isp's include for solid motors 280 - 300 seconds, for Oxygen and RP-1 350 seconds, for liquid hydrogen and liquid oxygen 450 seconds, and for ion engines 2000 - 6000 seconds [Sackheim, Wolf and Zafran, 1991], but ion engines have very low thrust compared to their engine mass. For comparison, NERVA, a simple nuclear rocket which forced hydrogen through a nuclear core where it was heated, was tested in the 1960s and provided an Isp of at least 925 seconds at high thrust [Borowski, 1990]. A proposed liquid core nuclear rocket, much more advanced but undemonstrated, would have high thrust and an estimated Isp of 2000 seconds [Ludewig, 1990].

Rocket's for interstellar missions appear to require significantly higher Isp's. A totally different architecture would be to propel a vehicle with a beam fired from a station in the departing location, which does not itself have to be accelerated. This avoids the need to accelerate reaction mass for later use, as a rocket does, and thus approximates an infinite Isp. Both particle beam and laser based versions of this architecture have been examined, and both appear promising [Andrews, 1994].

Another fundamentally different architecture than a rocket is using a cable to transfer momentum. The classic example is a skyhook, a cable in geosynchronous orbit long enough to reach down and attach to the ground. A payload which climbs the skyhook builds potential energy, gains momentum from the Earth's angular momentum, and can go into orbit. Another momentum transfer cable is a rotating skyhook, which can be arranged to have its tip velocity exactly cancel out the orbital speed and the Earth's rotation at its closest approach [Moravec, 1977]. Rotating cables can even be used in free space for trajectory modification [Mackenzie, 1995]. The feasibility of these concepts depends crucially on the ratio of tensile strength to mass density of the proposed cable materials.

2.2.4 Science Architectures

The Hubble Space Telescope, which engages in visible wavelength astronomy above the Earth's atmosphere, is an example of an architecture in this mission area.

2.2.5 Exploration Architectures

2.2.6 Exploitation Architectures

A classic architecture in this mission area is Solar Power Satellites (SPS's). They are large satellites in geosynchronous orbit, collecting solar power, producing electricity, and beaming that electricity at microwaves down to a rectenna assembly on Earth. Due to the beam diffraction at the desired wavelength and the geometry, and SPS like this must be roughly 5 GW or larger.

2.2.7 Enhance Quality of Life on Earth

2.2.8 Colonization

Another classic architecture rejects the idea that an advanced technological civilization necessarily belongs on the surface of a planet. [O'Neill, 1974a] was the first publication introducing O'Neill's particular system architecture of large, inhabited space colonies, with [O'Neill, 1974b] providing an overview of the details. The specific system architecture was an exercise in applied theoretical science. He took as requirements providing "normal gravity, normal day and night cycle, natural sunlight, an earthlike appearance, efficient use of solar power and of materials" [O'Neill, 1974b]. For this he proposed pairs of counter rotating cylinders, roughly 25 km long and 3.2 km in radius, divided lengthwise into three strips of habitable surface between equally sized window assemblies, with strip mirrors reflecting the sun into the colony. 72 much shorter agricultural cylinders would be associated with each cylinder. The assembly of two cylinders could support up to 13 million people in substantial comfort, or 20 million if the territory of the main cylinders were largely turned over to agriculture.

This example design supported a grand strategy where colonies could build more colonies, at a rate of up to once every 6 years, providing rapidly growing land areas, and thus the basis for much of humanity to leave Earth over the next century, bringing industry with them. The purpose of this grand strategy is to solve the following serious problems for humanity "without recourse to repression: bring every human being up to a living standard now enjoyed only by the most fortunate; protecting the biosphere from damage caused by transportation and industrial pollution; finding high quality living space for a world population that is doubling every 35 years; finding clean, practical energy sources preventing overload of Earth's heat balance" [O'Neill, 1974b]. Using resources from the asteroid belt, O'Neill estimated that even if growth rate could not be slowed, it could last for roughly 500 years, until the population had grown by a factor of roughly 20,000, and we would still have all the resources of the outer planets.

Despite this goal, and a process designed to be self-sustaining, it was not clear that Earth would provide the resources to start. [O'Neill, 1974b] had a series of progressively smaller colonies, with the ~13 million population version the largest "model 4," and the smallest a 10,000 inhabitant "Model 1," but this alone had an estimated cost of $30.7 billion, using 1972 dollars and assuming launch costs of no more than $425 per lb. O'Neill thus added a shorter term purpose of direct financial benefit to Earth--the colonies could produce SPS's which would beam power to Earth, and thus pay the costs of establishing the initial colonies [O'Neill, 1975]. Making a series of assumptions, including 10% interest charges, [O'Neill, 1975] estimated positive cash-flow in 15 to 20 years, and total payback in roughly 25 years.

Adding SPS's created O'Neill's basic operational concept. Build a Lunar Manufacturing Facility (LMF) on the Moon and Space Manufacturing Facility (SMF, a small colony with significant fabrication capability) in Earth orbit, for example at L-5. The LMF mines the moon, and launches material using a mass driver (linear accelerator) to the SMF, possibly using an intermediate spacecraft. The LMF and SMF production is used to build additional LMFs and SMFs, bootstrapping production. The SMF produces SPS's, which are then sold, or their power is sold, to utilities on Earth, which receive the power using rectenna farms. Later, the space infrastructure can be used to build larger space colonies. [O'Neill, 1976].

2.2.9 Planetary Defense

2.2.10 International Competition

The Apollo lunar vehicles themselves had as system architects Maxime A. Faget and Robert R. Gilruth [Rechtin, 1991]. An early driving question in the architecture was the mission profile. Obviously, the basic profile had to be go to the Moon, land, and come back, but there were a number of possible variations. Several in Huntsville wanted to dock two rockets in Earth orbit to make a super rocket, land that on the moon, and fly it back, but John Houbolt at Langley identified the mission profile ultimately selected--fly to lunar orbit, drop a lander leaving the return rocket in orbit, use the lander to return to lunar orbit, discard the lander, and return using a much smaller rocket than would have been needed to land on the moon, take off, and return [Gray, 1992].

After Apollo, this entire mission area lost much of its drive, as we had already "beat the Russians." With the end of the cold war, it currently has even less force.

2.2.11 Main Journals in the Field

The Journal of Spacecraft and Rockets, published by the AIAA, describes its scope as:

The Journal is devoted to reporting advancements in the science and technology associated with spacecraft and tactical and strategic missile systems including subsystems, applications, missions, environmental interactions, and space sciences. The Journal publishes original archival papers disclosing significant developments in spacecraft and missile configurations, re-entry devices, transatmospheric vehicles, systems and subsystem design and application, mission design and analysis,...space processing and manufacturing, space operations, interactions with spacecraft and sensors, design of sensors and experiments for space, and applications of space technologies to other fields....

Finally, Acta Astronautica, according to its aims and scope:

Acta Astronautica is the journal of the International Academy of Astronautics. It publishes original contributions in all fields of basic, engineering, life and social space sciences and of space technology related to: (1) the peaceful scientific exploration of space; (2) its exploitation for human welfare and progress; (3) the conception, design, development and operation of the systems, space-borne and Earth-based, needed to accomplish the foregoing two tasks.

2.3 Molecular Nanotechnology

What would happen if we could arrange the atoms one by one the way we want them?
• [Feymann, 1960]

Over time, technology is providing increasingly fine control over matter. We are manufacturing smaller features, as lines get thinner in computer chips, and micro-scale mechanisms are being produced [Mallon, 1992]. At the same time, technologies which already offer molecular precision are providing better control over larger products. Biotechnology produces atomically precise proteins, and progress in designing proteins [Sander et al, 1995], and in designing the sequence to produce proteins [Godzik, 1995], continues. Indeed, designed and published proteins have even gone through subsequent design iterations [Houbrechts et al, 1995]. Molecular modeling software is maturing, becoming interoperable between software packages, more capable, and taking advantage of the increasing cost-effectiveness of computers [Ahren, 1995], [Studt, 1995]. Scanning-Probe manipulation has gone from its most dramatic early demonstration [Eigler and Schweizer, 1990] to increasingly sophisticated atomic surface manipulations [Huang, Uchida and Aono, 1994]. Both trends, increasing precision with general manufacturing and increasing ability with atomically precise technologies, reach their logical conclusion in a general ability to fabricate with atomic precision.

Molecular Nanotechnology is the emerging ability to design and build systems to atomic precision. The idea of manipulating matter to atomic detail was foreseen [Feymann, 1960], and largely forgotten, years before it was unambiguously demonstrated [Eigler and Schweizer, 1990]. The key systems insight came independently in the 1970's to Dr. Drexler [Peterson, 1992], [Regis, 1995] when he realized that a manufacturing architecture based at the lowest level on molecular manipulations would be able to manipulate matter to atomic precision, that "The class of structures that can be synthesized by such methods is clearly very large," and that such production equipment would offer "the production of additional production equipment," making costs low [Drexler, 1981]. The natural thrust of molecular nanotechnology leads in the direction of the development and utilization of such molecular manufacturing systems [Feymann, 1960], [Drexler, 1986], [Drexler, Peterson & Pergamit, 1991], [Drexler, 1995a], [Merkle, 1995a], [Landman, 1995].

This future ability to fabricate at the natural limit of atomic resolution represents the digitization of matter. It will allow the processing of matter in ways analogous to digital processing of information, at the lowest level, using massive numbers of small, high-speed mechanisms. The result would be an ability to process matter with a speed, reliability, complexity, and flexibility similar to that which computers process information. If MNT meets its goals, its significance may even approach the significance of the information technology revolution.

The term "nanotechnology" has been widely used to refer to molecular nanotechnology [Drexler, 1986a], [Hall, 1993]. The term "nanotechnology" has also been widely used, however, to refer generically any technology involving objects with submicron features [Amato, 1991a]. The focus of this research is not on all submicron technology, but on those of wide molecular precision, allowing the manipulation of matter in a manner analogous to the digital manipulation of information. To avoid confusion, the term "molecular nanotechnology," and its abbreviation, MNT, are used here.

Since molecular nanotechnology will allow the fabrication of designed objects to atomic precision, product systems will be able to contain components with atomic-scale features, such as the design of a planetary gear in Figure 3 (next page), where the meshing of individual atoms forms the gear-teeth [Drexler, 1992]. MNT is to contain vast numbers of such simple molecular machines, combined into large, complex systems.

{Warning for HTML version of this file. If "2 x10⁵" looks like "2 x105" to you, then the exponents in the numbers below will not be correctly visible. I am generally using scientific notation. Do not be mislead by incorrectly formatted numbers. - Tom}

If an MNT system can have essentially every atom in its place, then the structural elements can consist of diamond, or even engineered substances such as a composite of diamond fibers in a diamond matrix (which would have greater fracture resistance than diamond). Such structures should have >5 x 10¹⁰ N/m² tensile strength, and a larger compressive strength. They would be ideal for many structural purposes. With a density of 3.51 g/cm³ [Pinneo, 1995], they would also have a strength to weight ratio of >1.4 x 10⁷ N*m/kg, which is 70 times better than aerospace aluminum (7178-T6, with a tensile strength to density of 2 x 10⁵ N*m/kg [Drexler, 1991]).

Figure 3. A molecular planetary gear, (a) end view; (b) side view; (c) exploded view. Figure from [Drexler, 1992a, p. 312] [This design is a collaboration of K.E. Drexler and R. Merkle.].

Based on conservative designs in [Drexler, 1992a], at least some MNT mechanisms should be capable of mechanochemical power conversion at a density of >10⁹ W/m³, while other MNT mechanisms should be capable of electromechanical power conversion at a density of >10¹⁵ W/m³. All of these mechanisms should be able to operate at power conversion efficiencies > 99%. Note that heat rejection problems would require volumes operating at these capabilities to be much smaller than one cubic meter, while design constraints may require individual devices to be at least on the order of 100 nm in size.

Also based on a conservative design in [Drexler, 1992a] built around molecular mechanical operations, some MNT-based logic gates should be as small as ~10^-8 u³ (~10^-26 m³). While meeting this size, it should also be possible for such gates to switch in ~0.1 ns, and dissipate <10^-21 J per operation. This should allow ~10¹⁰ MIPS per Watt. [Drexler, 1992a] adds an argument for cooling a ~10⁵ Watt, cubic cm system, at an equilibrium temperature of 300 K, which provides the capability for a compact, 10¹⁵ MIPS parallel computing system based on these characteristics. Note, since a major design driver was simplifying performance analysis [Drexler, 1988b], better designs with better performance are quite plausible. In particular, a nano-scale quantum electrical logic gate is likely to have a much shorter switching time. Whatever the nanocomputer design, they may become extremely inexpensive on a per-unit basis. If so, then they could be suffused throughout many systems.

The theoretical applied science argument in [Drexler, 1992a] culminated with an examination of molecular manufacturing. This is the process of building general products to atomic precision, using manufacturing systems which themselves can be presumed to have been built to atomic precision. This analysis indicated that a conservative unit operation rate for the lowest level devices would be 10⁶ operations/sec. The system analysis shows that it should be feasible to produce 1 kg of atomically precise general product, within 1 hour, using a molecular manufacturing system that itself masses no more than 1 kg. Note that in particular the molecular manufacturing system could build a copy of itself. This could allow swift amortization of the investment cost for the first molecular manufacturing system. Estimated marginal manufacturing costs for products were dominated by the cost of ordinary purity bulk chemical feed stocks, many of which could be taken directly from the atmosphere if necessary, and estimated in the $0.1/kg - $0.5/kg range.

Note however, that the exemplar architecture for molecular manufacturing used in Drexler, 1992a] was composed of ~3 x 10¹⁷ devices, each device potentially as complicated as a cascade of automated part feeders or a roboticized assembly cell. It should thus be clear that systems engineering is crucial to exploit MNT, particularly at the macroscopic level, and that many tremendously difficult problems remain to be solved before all of these performance estimates could be achieved.

Figure 4. If trend's in material processing continue, they lead to atomic precision.

There are several possible development paths for MNT. It is unclear which will succeed first, and thus prediction when MNT might be developed is uncertain. One can nonetheless guess. Current trends are no guarantee of future performance, but they are useful anchors to begin thinking about what might happen. If current trends continue, then indications are that MNT would be developed somewhere between 2010 and 2020 [Merkle, 1995b], as shown for example in figure 4 based on [Taniguchi, 1983]. This trend is not guaranteed, and should be taken as no more than a plausibility argument that MNT might be developed in a timeframe of interest.

Developing MNT may run into significant difficulties, delaying deployment. On the other hand, if developing MNT has significant first mover advantages, then when it gets closer there may be large, well-funded efforts to develop it. This dynamic might significantly shorten the final development period.

[Drexler, 1981] and [Drexler, 1986a] foresaw as a major milestone along one path to develop MNT the successful design and synthesis of de novo proteins. This was first achieved in 1987 [Regan and DeGrado, 1988].

A widely publicized step in the development of MNT was when researchers using a Scanning Tunneling Microscope (STM) manipulated 35 individual xenon atoms on a nickel surface, spelling out "IBM" [Eigler and Schweizer, 1990]. This was accomplished by cooling the sample to 4deg. K. Similar results were obtained at room temperature in later experiments [Huang, Uchida and Aono, 1994].

Dr. Drexler is the lead theoretician in the area, having published most of the classic, definitive, and influential pieces of research in MNT. After announcing the field in [Drexler, 1981], he publicized it in [Drexler, 1986a], further delimited it, especially with regards to public implications, in [Drexler, Peterson & Pergamit, 1991], and produced the seminal work in [Drexler, 1992a]. This work lays out the entire theoretical foundation, addressing a number of issues, showing how they can be considered, and that none are fatal. Issues addressed in [Drexler, 1992a] include thermal excitation, thermal and quantum positional uncertainty, quantum-mechanical tunneling, bond energies, strengths and stiffnesses, feasible chemical transformations, electric field effects, contact electrification, ionizing radiation damage, photochemical damage, thermomechanical damage, stray reactive molecules, device operational reliabilities, device operational lifetimes, energy dissipation mechanisms, inaccuracies in molecular mechanics models, limited scope of molecular mechanics models, limited scale of accurate quantal calculations, inaccuracy of semiempirical models, and providing adequate safety margins for modeling errors.

The only applications [Drexler, 1992a] discusses in any detail are computation and manufacturing (including intermediate technologies which would lead towards mature molecular manufacturing). A much wider array of applications have been discussed, however, with much less published quantitative justification. [Drexler, 1986a] discussed applications to manufacturing and material wealth, computers, space, health care (including radical procedures for currently general, incurable conditions), and aggressive and military uses. Applications in [Drexler, Peterson & Pergamit, 1991] include health care, environmental remediation (including atmospheric repair and species regeneration), armaments, education, entertainment, manufacturing, wall coverings, clothes, furniture, housing, food, communications, transportation, and safety from accidents and abuse. [Crandall, 1995] will discuss even more possibilities. The usenet group sci.nanotech acts as a giant slush pile, bringing forth even more suggestions, but most suggestions in this newsgroup have not been given critical review.

A prestigious work, winner of the Feymann prize in 1993, is [Musgrove et al, 1992], which analyzed a specific tool-mediated step, the removal of a hydrogen atom from a diamond surface.

The Foresight Institute has held three scientific conferences, specifically on MNT. The proceedings of the first conference evolved into [Crandall and Lewis, 1992]. The proceedings of the third conference resulted in [Krummenacker and Lewis, 1995]. The Fourth Foresight Conference on Nanotechnology will be November 9-11, 1995 in Palo Alto, and the proceedings will come out on the world-wide-web and in the journal Nanotechnology.

2.4 Systems Architectures for Space Operations Which Exploit Molecular Nanotechnology

As Drexler has a long-standing interest in space exploitation, which predates his thoughts in molecular nanotechnology [Peterson, 1992], [Regis, 1995], and he has been the seminal thinker in founding molecular nanotechnology, the bulk of important work combining these two fields has been his. Drexler's first book on molecular nanotechnology [Drexler, 1986a] devoted a full chapter at a popular level to the subject of space applications, discussing a fully flexible, one atmosphere, Closed Environment Life Support System (CELSS) space suit, asteroidal resources, interstellar, laser-driven lightsails, and the benefits of space resources and lower costs, all in the context of molecular nanotechnology. He also discusses building a rocket engine out of diamond in the book, and by implication entire space systems. The same year he published a paper [Drexler, 1986b] which primarily focuses on molecular nanotechnology, but concludes that this will allow producing "large structures made of nearly flawless materials," and that if these are spacecraft, then costs can be very low. In [Drexler, 1988a] he discusses specific space systems that could be built using molecular nanotechnology, although he does not address whether or not they would be operationally appropriate. The list is quoted below, but none of the numbers were substantiated in the paper:

• One-gee, 5,000 Isp solar electric propulsion system
• Synchronous skyhooks
• Thousand-kilometer diameter O'Neill-style space settlements
• Relativistic interstellar flight using laser-driven lightsails
• Turning a trillion tons of asteroid and a microgram of tools into a trillion tons of products in a matter of weeks.
  • [Drexler, 1988a]

In [Drexler, 1992b], numerically justified approximations are presented. He discusses a personal SSTO vehicle which delivers a 500 kg payload to LEO with a gross lift-off weight (GLOW) of ~3 tons, and a dry, empty mass of ~60 kg. He also discusses a 20 nm aluminum lightsail, and a solar electric ion engine with a 25,000 Isp.

Finally, [Drexler, 1995] also discusses applying molecular nanotechnology to space, but adds no worked out technical details beyond [Drexler, 1992b]. It does, however, recap a grand strategy for going into space--to utilize the extraterrestrial resources for a expanding technological civilization.

While appearing in the Journal of the British Interplanetary Society (JBIS), [Merkle, 1992] does not discuss space applications beyond making a general analogy in the introduction with proposals for self replicating interstellar probes [Freitas, 1980] and a self replicating lunar manufacturing facility [Freitas and Gilbreath, 1983], but instead focuses on the theory of self-replicating systems and molecular manufacturing as a particular, approaching implementation.

Also in JBIS, [Hansson, 1992] has some provocative ideas for space systems and operations, but it is not yet clear how to analyze them. [Amon, 1994] discusses using MNT for the information subsystem of a manned space vehicle, but the results are largely independent of this space context. Finally a number of articles in the nanotechnology issues of JBIS have focused on MNT instead of its applications in space [Peterson, 1992], [Kritsky and L'vov, 1992], [Santoli, 1992b], [Sheka et al, 1994], [Galushkin, 1994], [Santoli, 1994b], [Santoli, 1994c], or on other issues [Nussinov, Maron and Santoli, 1994], [Luskinovich, 1994].

James Bennet has long been both a member of the pro-space movement [Foresight, 1992], and long been an important official in molecular nanotechnology, serving on the Board of Directors of the Foresight Institute at least since 1998 [Foresight, 1988], serving on the Board of Directors of the Institute for Molecular Manufacturing (IMM) since its founding [IMM, 1991], and helped found the Center for Constitutional Issues in Technology [Foresight, 1992]. He is currently working on a popular book on space operations using molecular nanotechnology, but has not yet done much work analyzing actual capabilities [Bennet, 1995]. This research will be fed into James Bennet's effort, when and where appropriate.

[Toth-Fejel, 1984], in its discussion of self test for self-replicating robotic systems such as robotic lunar factories or molecular nanotechnology assemblers, raises an important architectural point. The problem of having confidence in base components is made much simpler by using atoms as the basic building blocks, since different atoms of the same isotope are utterly identical and fully interchangeable, simplifying quality control. It touches space applications of molecular nanotechnology, however, only as an aside.

The much more recent [Coppinger and Toth-Fejel, 1995], while still in progress, focuses exactly on the application of molecular nanotechnology to a mission in space. That mission is terraforming Mars. This is an excellent piece of work, well worked out quantitatively at the top level. Terraforming Mars using molecular nanotechnology is also the detailed focus of [Morgan, 1994] and sketched at in [Nussinov, Lysenko and Patrikeev, 1994].

Finally, [McKendree, 1995] lays out a pair of scenarios for space development that include operations based on molecular nanotechnology. This paper touches on architectures, operational concepts, and grand strategies.

There remains a gigantic hole in this body of work. For example, a search of "nanotechnology and space" on "Current Contents" in the USC periodical database finds nothing. Almost no systems architectures for space operations exploiting molecular nanotechnology have even been sketched at, and of those that have, only two provide even the barest numerical justification. There is a crying need to create a survey of the field, and to create at least representative entries across that survey. Meeting that need is the major focus of this research.

3.0 Research Objectives

These objectives are discussed in more detail in the subsections below.

It is also intended to clearly document the process used in order to establish a clear process of performing assessments of emerging technology to unprecedented systems, and to allow future researchers to study, critique, and follow as desired the steps involved.

3.1 Address Major Claims in the Literature About MNT and Space Operations

1a) Small, personal single-stage-to-orbit spacecraft [Drexler, 1988a],

1b) In particular, a ~60 kg dry mass vehicle able to deliver a 500 kg payload including four-passenger with luggage and life support to Low-Earth Orbit, with a gross lift-off mass of ~3 tons [Drexler, 1992b] (some calculations shown).

2) A fully flexible, one atmosphere, Closed Environment Life Support System (CELSS) space suit [Drexler, 1986a].

3) A mission profile in which individual citizens settle and survive in space, using only the financial resources available to a typical first-world citizen [McKendree, 1995].

4) One-gee, 5,000 Isp solar electric propulsion system [Drexler, 1988a].

5) A ~0.8 m/s² solar electric propulsion system with a ~25,000 Isp [Drexler, 1992b].

6) Synchronous skyhooks [Drexler, 1988a].

7) Thousand-kilometer diameter O'Neill-style space settlements [Drexler, 1988a].

8) Relativistic interstellar flight using laser-driven lightsails [Drexler, 1986a], [Drexler, 1988a].

9a) Utilization of self-replicating assemblers to convert asteroidal material into useful product [Drexler, 1986a],

9b) In particular, turning a trillion tons of asteroid and a microgram of tools into a trillion tons of products in a matter of weeks. [Drexler, 1988a].

3.2 Performance on Space Operations Functions

1. Earth to Earth-orbit transport
2. Earth orbital transfer
3. Payload return to Earth
4. Signal relay
5. Docking payloads with low-gravity objects
6. Landing payloads on hard, moderate gravity bodies (both with and without atmospheres)
7. Remote sensing of Earth
8. Astronomical observation
9. Remote sensing of extraterrestrial bodies
10. Life support in a hostile environment
11. Micro-g research facility
12. In situ science
13. Extraterrestrial maintenance and repair
14. Extraterrestrial resource recovery
15. Interplanetary transport
16. Intercept of threatening bodies
17. Interstellar transport.

3.3 Generate a Novel Space Systems Architecture

Note, given an architecture independently developed in this research, the claim of originality requires a second literature search, specifically for a prior mention of an equivalent architecture. Those searches will be part of task 7.

4.0 Planned Research Activities

Figure 5. The research approach provides the necessary integration of system architecture, operational concepts, and grand strategy into consideration of an MNT-based space systems.

The research is decomposed into 8 tasks. The flow of these tasks leads the research through a complete loop from system architectures through operational concepts to grand strategies, and back. This flow, shown in figure 5, embeds into the larger long-term process shown in figure 6.

Section 4.1 below discusses why theoretical applied science is an appropriate method for the research objectives, and why the tasks have been decomposed and arranged as shown in figure 5. Section 4.2 discusses the skills needed to perform this research. Section 4.3 discusses each of the tasks which make up this research. Section 4.4 discusses the current schedule for completing this research. Finally, section 4.5 discusses various risks inherent in this research, and where and how those risks are being mitigated.

4.1 Justification for Chosen Method

4.1.1 Justification for Theoretical Applied Science

Propositions in theoretical applied science are falsifiable if one adopts the rule that any conflict with established scientific knowledge constitutes falsification.
• [Drexler, 1992a]

The three pillars of this research are system architecture, space operations, and molecular nanotechnology. While many system architects are currently practicing, and many system architectures and space operations exist today, mature MNT unfortunately does not currently exist in a physically testable form. This makes the research theoretical rather than experimental. Space systems are meant to be useful devices expressing intended purposes. This makes the research applied. The combination makes this research an exercise in theoretical applied science.

Table 2. Theoretical Applied Science fits among the dichotomies of theoretical versus experimental and pure versus applied research. Derived from [Drexler, 1992a].

                            Pure                            
Applied

Theoretical  Pure Theoretical Science: What is    Theoretical Applied
Science: What    
               the description of a fundamental     is a minimum estimate for
the        
               physical law obeyed in the           maximum performance a
class of       
               universe?                            useful devices can
achieve?          

Experimental Pure Experimental Science: What      Experimental Applied
Science:        
               are measured observations of         Design and build a device
in         
               actual events in the physical        accordance with scientific
laws.     
               universe?                            Does it work?  Is it
useful?         

Drexler argues that because scientists usually focus on questions that are informed by, or can be addressed with, existing or near-term experimental devices, and engineers usually focus on products that can be profitably produced with current or near-term tools, theoretical applied science is often overlooked [Drexler, 1992a]. One can argue, however, that many concept analyses and feasibility studies are examples of theoretical applied science. Indeed, the early stages of system architecting for novel systems, when concepts are still to be reduced to hardware, is very much in the same spirit. Theoretical applied science results also can be important early data for systems architecting efforts. Some theoretical applied science research differs from early systems architecting activities, such as concept analysis and feasibility studies, in that the latter generally include a requirement that the system be physically realizable with existing or near-term tools.

[Drexler, 1992a] describes how, when the goal in theoretical applied science is to set minimum estimates on the maximum achievable performance of classes of devices, the analysis need not produce a design which is manufacturable with existing or near-term tools, nor enough detail to manufacture a device. This provides some flexibility which can compensate for approximations in the models and the inability to directly experiment. Indeed, with this goal there is no need to attempt to create designs which are likely to actually be built. In the near term, the tools may not exist, and in the far term, better designs are likely to be available.

In the research being proposed, the goal is to understand the intrinsic capabilities of MNT-based space systems. It will do so by developing minimum estimates on the maximum performances from this class of systems, and relating those estimates to performance capabilities in specific operations.

Since including the subject of molecular nanotechnology requires this research to focus on devices which cannot be soon built, this provides the opportunity for the research to also examine classes of space systems which cannot soon be built. Indeed, space system architectures may require MNT to be feasible, while offering significant benefits. This research specifically invites such system architectures.

An important research study that has been successfully carried out using the theoretical applied science examination of minimum estimates on the maximum performance possible from defined classes of devices is [Drexler, 1992a].

4.1.2 Justification for Task

The task flow is shown in figure 5 above. Tasks 2 - 4 are the synthesis of MNT with previously defined concepts, and a bottom-up analysis of the resulting system architectures, operational concepts and grand strategies. Tasks 5 - 7 conduct top-down definition and synthesis, informed by the earlier tasks. The literature search in task 1 is to provide the basis of previously defined system architectures, operational concepts and grand strategies, along with their performance models, to support tasks 2 - 4. Task 8 compiles the results into a publishable dissertation. This task flow embeds into a larger process, discussed in 4.1.2.1 below. Other possible research task flows are discussed and compared in 4.1.2.2 below.

4.1.2.1 The Larger Process

Figure 6 on the next page illustrates an iterative approach, integrating the technical capabilities of MNT with their implications for systems, operational concepts, and grand strategy. This approach should evolve answers which converge on an increasingly accurate assessment of MNT's implications for space operations.

The iterative process in figure 6 never ends, it merely keeps refining the answer. This process will likely to go on for years or decades, in the efforts of many researchers, through the development of MNT and space applications. For purposes of this dissertation, the process will be taken through one full loop from space systems through grand strategy, and back. The process will start with previously analyzed system architectures, operational concepts, and grand strategies, to provide a firm grounding. This firm grounding is crucial for confidence, as initial efforts are likely to include such counterintuitive results as a rocket which can deliver 500 kg to low earth orbit, but which has a dry empty mass of only ~60 kg [Drexler, 1992b].

Figure 6. An iterative process for assessing system architectures using MNT, likely to continue for years or decades by many researchers across many research projects. The proposed research embeds into this process.

Once this grounding is established, the process will then return to each of these areas in turn, grand strategies, operational concepts and system architectures, to particularly exploit the apparent strengths and features of MNT. The step is crucial, as MNT offers significant novel functions and significantly improved technical performance, undoubtedly requiring novel architectures to fully exploit.

4.1.2.2 Comparison with Other Task Flow

A common advantage of both approaches is that being more direct they should be easier and faster. The first approach has an issue with the objective of developing a supported minimum estimate for the maximum performance for major space system function using a system architecture with MNT. When directly looking at each function, there is a risk that a sufficiently wide range of system architectures may not be considered. Furthermore, if the nature of MNT is such that it will provide for a novel major space system function, going through a check-off list will not discover that function.

The first approach also runs into potential difficulties when it gets to the end, if it has not already generated at least one novel system architecture for space operations which is realizable within our understanding of physical law. Since it is narrow in exploration, it may have difficulty moving sufficiently off the beaten path to find a novel architecture.

The second approach makes the process more direct by narrowing the scope down to a single grand strategy, along with its related operational concepts and system architectures. The second approach does maintain, however, a research objective of understanding in some detail the possibilities for systems to implement the grand strategy selected. It is not likely to examine the full range of major space system functions, nor to notice if MNT is better suited to some other grand strategy. It does have a good chance of discovering any novel major space system function that may exist within the purview of the selected grand strategy. There is some danger that the second approach would lack sufficient generality to be truly useful, but since the process of top-down systems engineering should include adequate trade studies to make conclusions across ranges of possible systems, the results should be quite general within the context of the selected grand strategy.

The selected approach has the difficulty that it is larger in scope. It will provide analysis of a wider array of system concepts using MNT. The process has more scope to understand what sorts of operational concepts and grand strategies are more preferred in the context of MNT, and to seek system architectures which better harness MNT's technical capabilities to system performance. A weakness is that the set of generated concepts is likely to be somewhat idiosyncratic--a different researcher would probably generate a different set.

4.2 Required Research Skills

Table 3. Skills needed for the proposed research.

Needed Skill                     Tasks         Skill Acquisition Strategy  
             

Library Research                 1, some 7,    Largely acquired.  Working with
Julie     
                                 8, maybe 2,   Kwan in the Science and
Engineering       
                                 3, 4          Library.                       
          

Concept Analysis                 2, 7          Already developed              
          

Ability to use already defined   2, 3,         Already developed              
          
mathematical models                                                           
          

Ability to write executable      2, 3, maybe   Already developed              
          
computer simulations, given a    4                                            
          
description of the model                                                      
          

Mathematical modeling            2,3,6,7       Already developed              
          
                                 maybe 4, 5                                   
          

Computer modeling and            6, 7, maybe   Already developed              
          
simulation                       3, 4, 5                                      
          

Space system design              7, some 6,    Largely acquired.  Further
developing     
                                 maybe 2, 3    through practice               
          

Operational concept analysis     3, 6          Already developed              
          

Strategic Analysis               4, 5, some    Already developed (see for
example,       
                                 6, maybe 3    [McKendree, 1988])             
          

Strategic synthesis              5, some 4,    Already developed              
          
                                 maybe 3, 6                                   
          

General writing                                Already developed              
          

Dissertation writing             8             Draw on general writing skills.
 Review   
                                               other dissertations.  Attend
the          
                                               Graduate School's dissertation
writing    
                                               short course                   
          

4.3 Research Tasks

Figure 7. The examination of MNT-based space systems architectures, and to a lesser extent, operational concepts, directly supports the research objectives.

These tasks are described in more detail below. As part of the task descriptions, two examples are carried through the sequence of steps, orbital skyhooks and self-replicating lunar mining/factory complexes (SRS's). The literature survey is the first step, and thus has the most detailed description.

4.3.1 Literature Survey

Likely sources are the Journal of the British Interplanetary Society, Acta Astronautica, the Journal of Spacecraft and Rockets, AIAA Papers, and the proceedings of the Biannual Space Manufacturing Conferences held by the Space Studies Institute.

For Earth to Earth-orbit transport architectures, specific items to look for are stacked multiple stage disposable rockets, the semi-reusable 1 1/2 stage shuttle, reusable single-stage to orbit (SSTO) concepts, and skyhooks (aka "beanstalks," aka "orbital towers"). For skyhooks, [Pearson, J, 1975] "The Orbital Tower: A Spacecraft Launcher Using the Earth's Rotational Energy," Acta Astronautica, September-October 1975, is the foundational citation. The first Western publication of this idea was [Isaacs, J.D., H. Bradner, G.E. Backus and A.C. Vine, 1966] "Satellite Elongation into a True 'Sky-Hook,'" Science, 11 February 1966. This research has a copy of [Moravec, 1977] "A Non-Synchronous Orbital Skyhook," which discusses a different architecture for very long rotating cables in space, which are not fixed to the ground.

Specific architectures to look at for Earth orbital transfer include solid booster rockets, refuelable "orbital tugs," and Rotating tethers (aka "bootlaces" [Mackenzie, 1995]).

Specific architectures to consider for payload return to Earth include simple re-entry capsules like the Apollo Command module, aeromaneuverable reentry devices such as the Space Shuttle, aerobrakes, vertical burn landers such as the Delta Clipper, parachutes, and skyhooks. In looking for landing payloads on hard, moderate gravity bodies (with and without atmospheres), the specific architectures for payload return to Earth are candidates.

Geosynchronous communications satellites, LEO communications constellations and interstellar radio beacons and radio telescopes are at least three signal relay architectures to look for in the literature.

For docking payloads with low-gravity objects, specific architectures to look for include maneuvering retrorockets, cable hookup concepts, and the non-docking architecture of close co-orbiting.

In looking for remote sensing of Earth, declassified information on intelligence satellites, and information on remote sensing satellites, like LandSat, are specific subjects to look for in this literature survey.

Specific architectures for astronomical observation include Earth-orbiting telescopes like Hubble, and proposals such as for a far-side lunar astronomical observatory. They include not only systems intended to look deep into the universe, but telescopes searching for orbital bodies within the solar system.

In looking for specific architectures for remote sensing of extraterrestrial bodies, deep space science probes such as those produced at JPL are the classic examples.

In looking for life support in a hostile environment, previous manned missions provide one architecture. The Space Station offers a refinement of that architecture. Fully closed environmental life support systems built around creating a biosphere offers another specific architecture. Finally, proposals to use MNT to terraform other planets are an extreme life support architecture in the literature.

Micro-g research facility architectures to look for include manned LEO stations (such as Mir and Skylab), and unmanned free-floaters.

System architectures for in situ science are likely to be tied to specific mission location concepts. Places to look include specific proposals for Lunar and Mars bases, automated concepts such as the Galileo atmosphere probe and planetary landers, and even rendezvous transponders proposed for rapid orbit determination.

System architectures for extraterrestrial resource recovery are also likely to be tied to specific resource locations, and to specific resources. Examples include Lunar mines, Asteroid mines (with metallic asteroid mining and carbonaceous asteroid mining likely to be very different), Lunar He³ mines, and Solar Power Satellites. An area in which a pre-existing architecture needs to be examined is bootstrapping (including self-replicating) systems for utilizing extraterrestrial resources. [Freitas and Gilbreath, 1983] discusses self-replicating systems on the Moon and appears to be the major work in this area.

System architectures for interplanetary transport and trajectory modification primarily vary on the propulsion system. Specific examples to look for include chemical rockets, Solar Sails, nuclear-powered ion drives, solar-powered ion drives, unsymmetric concentrated solar illumination, and momentum transfer via long tethers.

Architectures for intercepting threatening bodies are addressed in works which examine planetary defense more broadly. Hazards due to Comets and Asteroids, edited by Tom Gehrels, University of Arizona Press, 1995, appears to be the definitive work on the subject, and is a major target of the literature survey. Another source is The Threat of Large Earth-Orbit Crossing Asteroids, Hearing before the Committee on Science, Space and Technology; US House of Representatives, March 24, 1993, available from the Government Printing Office.

The Orion concept, the Bussard Ramjet, lean Anti-hydrogen+hydrogen plasma rockets, laser-powered light-sail (with "Forward" brake), and Magsails are all specific architectures for interstellar transport to look for in the literature survey.

In the literature survey, I will also want to search for any concepts specifically thought to have applicability to intergalactic transport. This might include any of the interstellar architectures.

Another major purpose of the literature survey is to locate existing and proposed operational concepts for space operations. Examples clearly include commercial communications satellites, the Apollo program, suggestions for a space hotel, the "Mars Direct" proposal for missions to Mars, Gerard K. O'Neill's proposal [O'Neill, G.K., 1976] for a "system of systems" including lunar mine and factory, High Earth Orbit Colonies and factories, and Solar Power Satellites, NEO detection and intercept (before Earth impact), and NEO exploitation missions.

A third purpose of the literature survey is to locate existing and proposed grand strategies for space operations. Two examples to specifically look up include the vision of the Planetary Society, which appears to be to study space for its scientific aspects, the vision of the old L-5 Society and current Space Frontier Foundation that space should be a place for free settlers in the spirit of the American West. A related grand strategy is to develop planetary defense against Earth crossing bodies, and grow that capability into NEO exploitation and ultimately full utilization of the asteroid belt. Also, the International Space University's Summer 1995 Session had a project looking at suggested space plans for 25 years, Vision 2020. The literature search should get this document if it becomes available in time.

Another important subtask will be to monitor the literature to make sure this research has the most current estimates of MNT performance.

A related target is to locate currently unidentified concepts for MNT-based space systems, operations, or grand strategies. While everything by Drexler and McKendree are located, sci.nanotech represents a large "slush pile" of possible concepts. It is archived at planchet.rutgers.edu on the internet.

There will also need to be some research in the literature to look up standard data on the space environment. Issues include the radiation background, both ambient and exceptional, and the distribution of elements on heavenly bodies.

The reference list for this research proposal provides a starting point for the literature survey. There is a comparable set of citations already identified for this research which are not cited in this proposal, and thus not shown in the reference list. They are primarily papers on various space system architectures defined without MNT, which remain to be read.

4.3.2 Assess Space System Architectures in the Literature, Using the Technical Parameters of MNT

This task will rely on pre-existing performance models, and some models will not be in the literature. Furthermore, the various performance models make different assumptions, and for some models the MNT technical performance parameters will be outside the range where the assumptions made for the model are valid, even after making minor modifications to the models. This research will preferentially select system architectures in the literature for which performance models are available which can properly use MNT parameters. To indicate the degree of selection bias this approach may have introduced, the research will note the fraction of system architectures not selected due to lack of a convenient performance model, out of the total number of system architectures looked at in the literature survey.

The assessment for skyhooks would be to use the best possible strength-to-density material ratio from MNT in the equations in [Pearson, 1975] and [Moravec, 1977], and compare that to the performance they estimated achieving.

For example, [Moravec, 1977] analyzed the rotating skyhook system architecture. One conclusion was that a orbital rotating skyhook with a length equal to 1/3 the radius of Earth is a good architecture, because it is near the minimum mass, and can be arranged to make its closest approach repeatedly at six stable locations on Earth. That paper derived an for the taper ratio, the ratio between the largest area and the smallest area of the cable, assuming a circular orbit and a tip speed that just cancels out the orbital velocity and Earth's rotation when it makes its closest approach.

Assuming graphite crystals, [Moravec, 1977] derived a taper ratio for a 1/3 Earth radius skyhook of 10.05. Using the same equation, a density of 3.51 g/cm³ and the tensile strength of 5 x 10¹⁰ N/m² from section 2.2 above for the cable, yields a 5.5 taper ratio for an MNT-based 1/3 Earth radius skyhook.

[Freitas and Gilbreath, 1983] provides a moderately well-defined logical architecture for a lunar SRS. This research will go though this architecture, looking for parts which could be implemented using MNT, and asses the overall system performance assuming those components use MNT. Equations for overall system performance are not defined in [Freitas and Gilbreath, 1983] down to subsystem parameters, and would have to be derived.

For this particular case, there may be a somewhat difficult architecture issue. On the one hand, most current estimates for improved performance of MNT assume carbon-rich devices. On the other hand, the lunar maria is only 104 parts per million carbon [Freitas and Gilbreath, 1983]. One can forgo most beneficial MNT devices, or one is forced to an overall SRS that must extract and process noticeable amounts of carbon from a very thin lunar environment. This would require a trade study to balance. The most useful applications in this architecture for MNT may be in producing low-mass fraction components such as computer chips, sensors and ball-bearings which otherwise might have to be imported from Earth.

It is hoped that one result of this task will be to provide a general feel for what classes of space systems and purposes MNT is particularly well suited for.

As an example of something cited in the literature, for which published quantitative justification remains, is the "Thousand-kilometer diameter O'Neill-style space settlements" from [Drexler, 1988]. This requires merely showing calculations based on the basic equations for O'Neill-style space settlements, using the estimated technical performance parameters of MNT. It is not clear that this would have great utility

As a general rule, architectures assessed in this step are anticipated to show increased performance when using the technical performance parameters estimated for MNT, but are not anticipated to be well matched to MNT. Thus, these architectures may fall well short of what space systems designed and balanced to exploit MNT would achieve. Furthermore, many of these architectures may be unbalanced in that they offer some spectacular capability with is of no operational utility. Responding to the former issue is the purpose of task 7, and assessing the latter issue is the purpose of the next task.

4.3.3 Assess Concepts For Space Operations in the Literature, Using the Technical Parameters of the MNT-Based Space System

For example, the performance of a rocket system architecture would include the thrust capability and the total [[Delta]]V (capability to change velocity). An operational concept might be to take a payload to Mars. Other concepts of operation may be more complex. For example, orbital telescopes monitor for approaching comets. When detected, a high-speed probe brings a rugged transponder to the comet, allowing range and range-rate determination by radar. The trajectory is determined, and if it is threatening, probes in space using pre-positioned energy stores rendezvous with the comet and deflect it to a safe trajectory.

The [Freitas and Gilbreath, 1983] SRS could be used in the [O'Neill, 1976] operation concept : 1) set up mining and processing on the Moon; 2) set up manufacturing in space supported by space colonies; 3) using resources mined and shipped from the Moon the space manufacturing builds Solar Power Satellites (SPS's) and sells them to Earth, with the intent of earning enough to cover the costs for all the previous steps. The performance of this concept for space operations, using the performance estimated in the task above for the MNT-exploiting SRS, would be calculated in this task.

[Merkle, 1990] observed about self-replication "A significant advantage of this approach for space exploration would be to reduce or eliminate the need to transport mass from Earth--which is relatively expensive," and it is avoiding that cost that drove consideration of the SRS concept. With an MNT-based skyhook, however, transporting mass from the Earth might not be expensive. This points to a second operational concept: 1) Transport vast quantities of equipment and material into space. 2) Build SPS's, and sell the power back to Earth. With the reduced cost of the putting material in space, due to the skyhook, maybe this operation concept becomes superior. This second operational concept would be analyzed using the Earth-launch system that was estimated as most favorable in the previous task.

As a final example, [Friedman, 1995a] suggests "seeds in space," unmanned precursor probes with low mass and high performance to build up operations in space the preparing way for ultimate colonization missions, and also suggests "Oceans of Energy," creating many caches of separated hydrogen and oxygen, available as pre-processed resources and fuel stations. MNT appears ideally suited for the seeds in space operational concept, and may better support the oceans of energy. On the other hand, MNT may provide robust solar sail and momentum transfer systems, reducing the operational need for "oceans of energy," at least for propellant.

This task should provide a sense of what sorts of missions MNT is particularly well suited for, a basis of results to support the next task, and hopefully some intuition which will assist in tasks 5 and 6.

4.3.4 Assess Grand Strategies For Space Operations in the Literature, Using the Estimated Performance of the MNT-Based Operational Concepts

Grand strategies can be compared in at least the following two senses. First, are the aims and objectives of one grand strategy better achieved when following another grand strategy? Second, if two grand strategies were each pursued by different capable actors, with whatever interference that implied, how well would the grand strategies achieve their respective aims and objectives? Quantitative measures of effectiveness (MOEs) which are general across grand strategies are not expected, but MOEs within grand strategies are plausible, and operational concepts in support of grand strategies are expected to have well defined MOEs.

Every grand strategy for operations in space involves some amount of launching material from earth, and thus the estimated capabilities of the skyhook may impact any grand strategy. If the skyhook makes affordable launching tremendous amounts of material into orbit, that is likely to favor the grand strategies which make more use of Earth-launching vast supplies.

The lunar SRS only matters for those grand strategies with substantial processing of lunar material.

Hopefully this task will provide the guidance for developing any major grand strategies well suited to MNT that are not in the literature. This thought about grand strategies, in conjunction with task 5, is needed to support task 6.

4.3.5 Generate and Assess MNT-Based Grand Strategies

This task will be to generate grand strategies specifically inspired by the estimated capabilities of MNT space systems, and assess those strategies in the same manner as in the previous task. These new grand strategies, if any, are likely to be inspired by the grand strategies, and their strengths or deficiencies, from the previous task. The purpose of this and the previous task are to provide context for the operational concepts developed in the next step.

Since quantitative figures of merit between grand strategies are not expected, tasks 4 and 5 may yield a subset of multiple conflicting grand strategies where none is clearly superior. If so, then an attempt will be made to design operational concepts and system architectures in the next two steps so that the performance of those concepts and architectures will help indicate which of these grand strategies is superior.

If the skyhook offered a tremendous increase in capabilities, that might suggest developing a grand strategy centered around deploying and using the skyhook. Similarly, a truly tremendous result from the analysis of the lunar SRS might suggest a novel grand strategy (although the basic grand strategies built around lunar exploitation seem to have already been identified). More generally, the earlier tasks, including the assessment of the lunar SRS and skyhooks, may provide a sense of what MNT is particularly well suited for, which could suggest a novel grand strategy playing to those strengths.

In the course of the previous tasks, ideas may come up for MNT-based grand strategies. If so, those ideas will be captured when they occur, and in this sense, this task will begin at the beginning of the research. Also, the assessments in this task may depend on the performance of operational concepts and system architectures to be developed in the next two tasks. In those cases, final conclusions will remain open until the subsequent performance calculations are performed, and in that sense, task 5 will not be completed until after tasks 6 and 7.

4.3.6 Generate and Assess MNT-Based Concepts For Space Operations

For constant mass, smaller satellites allow more satellites per constellation. MNT might allow very small satellites, and thus huge constellations. Just because the constellation can be huge does not mean that small satellites is the correct architecture for a system.

Consider for example a constellation of sun-orbiting satellites with large on-board telescopes searching for new comets entering the solar system, in order to provide early detection and warning. Assume a constant constellation mass, and that l is a characteristic length on individual satellites. The volume of space to be surveilled is constant. Assuming constant thickness telescopes (balloon Cassigrains) gives satellite masses that scale approximately as l². The telescope diameters will scale as l. Detection range (for constant brightness comet) scales as l. The volume surveilled per satellite (for a constant field of view) scales as (Detection range)³, so it scales as l³. Therefore, for constant constellation mass, volume survielled scales as l³/l². or l^3/2. Thus, larger characteristic length (l) is better, and a smaller number of large satellites is better than a huge number of tiny satellites. Note, the constant thickness assumption may be too conservative. The total system mass should not scale any worse, however, than l³. In that case there is no dominant mass penalty (or mass benefit) to scaling.

An operational concept which looks favorable at this very gross level of analysis would deserve a much more detailed examination.

For another example, the standard concept for constructing a skyhook is to establish a geostationary satellite, and then build outward from that until the skyhook is essentially complete, supplying all the mass in orbit. Another approach would be to first build a very thin skyhook. Then, using MNT, the skyhook could draw carbon out of the atmosphere, and use that as the mass to build up the skyhook until it is fully constructed.

Also, a diamondoid structure will degrade at some rate due to background radiation in space. This would be particularly pronounced around the Van Allen radiation belt. Small MNT devices could run up and down the cable, repairing radiation damage at the atomic level. This concept must also include some means to deal with radiation damage to the MNT devices themselves.

For third example, while the lunar SRS is meant to operate on the moon, it is not able to exploit as much of the currently estimated performance of MNT, because those estimates are for carbon-rich devices, and the moon is very carbon poor. An alternate operational concept would be to operate in a carbon rich location. This gives rise to the operational concept of placing some SRS on a carbonaceous asteroid. A new requirement for such a system is being able to do all operations in the microgravity environment of the asteroid docking.

The "Sooners" scenario in [McKendree, 1995] presented another MNT-based concept for space operations: 1) Covertly build and launch a rocket that delivers an SRS on an asteroid {this operation concept assumes the SRS can be made smaller than 1 kg}. 2) The SRS and its progeny take over the resources of the asteroid, turning it into useful equipment, including a desirable place to live. 3) In parallel, covertly build, and then launch a personal vehicle that takes oneself and close others to the asteroid. 4) Dock with and claim the asteroid and all its resources. Assessing this operational concept requires developing and assessing the component system architectures.

In the course of tasks 1 through 5, ideas may come up for MNT-based operational concepts. If so, those ideas will be captured when they occur, and in this sense, task 6 will begin at the beginning of the research. Also, the assessments in this task may depend on the performance of system architectures to be developed in the next task. In those cases, final conclusions will remain open until the subsequent performance calculations are performed, and in this sense, task 6 will not be completed until after task 7.

4.3.7 Generate and Assess MNT-Based Architectures For Space Systems

Several papers have been published on "nanosatellites" [Janson, Helvajian and Robinson, 1993], [Janson, 1994]. This term is used, however, to refer to satellites on the order of 10 cm across, with a mass of ~1/3 kg. An open area of exploration for this task is "true" nanosatellites, presumably massing closer to a picogram, and requiring MNT to produce, although, they would require some useful mission to have been found in task 6 to be worth pursuing.

This task could also analyze a skyhook that drew carbon out of the upper atmosphere to grow. How much surface area is available in the atmosphere? What is the carbon uptake rate? How is the mass redistributed along the length of the cable, and at what performance? How fast can the cable grow itself? Note, this idea may prove greatly inferior to simply returning adequate carbon for a full skyhook from extraterrestrial sources to geosynchronous orbit.

This task would also include designing at the top level small MNT devices that move along a skyhook, repairing radiation damage. Their system architecture would include describing in quantitative detail how they repair the radiation. Analysis would have to include how effective the repair effort would be, and how a population of these devices would continue to operate in the face of the radiation environment.

An issue identified in [McKendree, 1995] as critical to MNT and space is the design and performance of an SRS intended to operate on a carbonaceous asteroid. This requires creating the top-level design, perhaps in parallel detail to the [Freitas and Gilbreath, 1983] lunar SRS, of this asteroid SRS. It also requires estimating the basic system parameters, such as initial mass, doubling period, allowable asteroid inhomogenaity, and parts list size.

The earth-to-interplanetary, small ("backyard") launch vehicle was already sized as background work [McKendree, 1995]. These calculations would be revisited in light of the work done in this research, particularly the estimated initial mass of the carbonaceous asteroid SRS.

The personal interplanetary vehicle was also already sized as background work for the SSI paper, and those calculations would also be revisited in light of the work done in this dissertation.

The third research objective, generate at least one novel system architecture for space operations which is realizable within our understanding of physical law, will be achieved in task 7. The key to inventing a novel systems architecture is likely to be in exploiting a novel function which MNT will enable, such as on-board self repair, or radical self reconfiguration under remote or automated control. It might also be possible with MNT to create an architecture with multiple, loosely coupled, self-replacing subsystems, such that the entire system architecture resembles a full, self-contained biosphere. Such a system might adapt across a wide range by changing the population distribution of its various subsystems.

During the previous tasks, ideas may come up for MNT-based space systems. If so, those ideas will be captured when they occur, and in that sense, this task will start with the beginning of the research.

4.3.8 Write Dissertation

The currently foreseen outline of the dissertation is shown below:

Front Matter

1.	Introduction
2.	Background
2.1	Systems Architecture
2.2	Architectures for Space Operations
2.3	Molecular Nanotechnology
2.4	Systems Architectures for Space Operations Which Exploit Molecular
	Nanotechnology
3.	Documentation of the Process
4.	Grand Strategies for Space Operations
5.	Operational Concepts for Space Operations
6.	System Architectures for Space Operations
6.1	Architectures for Transport from Earth
6.2	Architectures for Earth Orbital Change
6.3	Architectures for Payload Return to Earth
6.4	Architectures for Signal Relay
6.5	Architectures for Docking Payloads with Low Gravity Objects
6.6	Architectures for Landing Payloads on Hard, Moderate Gravity
	Bodies
6.7	Architectures for Remote Sensing of Earth
6.8	Architectures for Astronomical Observation
6.9	Architectures for Remote Sensing of Extraterrestrial Bodies
6.10	Architectures for Life Support in a Hostile Environment
6.11	Architectures for Micro-Gravity Research
6.12	Architectures for In Situ Science
6.13	Architectures for Extraterrestrial Maintenance and Repair
6.14	Architectures for Extraterrestrial Resource Recovery
6.15	Architectures for Interplanetary Transport
6.16	Architectures for Intercept of Threatening Bodies
6.17	Architectures for Interstellar Transport
7.	Conclusions

Appendices
Bibliography

Figure 8. Each major analysis task feeds into a specific section of the dissertation.

of the architecture with molecular nanotechnology. System architectures which could be classified in more than one section of chapter 6 will be placed where they have the greatest affinity, and cited from the other sections.

Chapters 5 and 4 should have subordinate organization, but the details remains unclear.

>From a readability standpoint, the order of chapters 4, 5 and 6 could be reversed. This change is still being considered.

Section 1.0 & 2.0 of this research proposal serve as preliminary rough drafts for the Introduction and Background chapters of the dissertation, with section 3.0 of this proposal also contributing to the Introduction. Section 4.0, serves as the initial input to the Documentation of the Process chapter of the dissertation. Section 4.3.2 and 4.3.7 include early contributions to the Systems Architectures for Space Operations chapter, and sections 4.3.3 and 4.3.6 include early contributions to the Operational Concepts for Space Operations chapter. The references will carry over into the bibliography.

As figure 8 shows, the various tasks of the research directly feed into particular portions of the dissertation outline. This will aid in maintaining a rough draft of the dissertation in parallel with the research.

An idea to consider is maintaining a draft versions of the dissertation work in progress on the World Wide Web, inviting comment and criticism. Significant comment without telling criticism would provide some additional confidence in the results. This would only be done if there is no danger of creating priority difficulties or other problems with the dissertation itself.

4.4 Current Schedule

Figure 9. Current schedule for dissertation research.

4.5 Risks and Mitigation

The first risk comes from the breadth of the research effort, that it is of excessive scope to complete. The mitigation plan is implicit in the nature of the risk. If the scope is too large, then it logically follows that the scope of this research is larger than necessary, and a restriction of scope would be permissible. There are several dimensions along which restricting the scope is feasible. The best might be to select a single grand strategy, which then becomes the theme of the dissertation and of all the operational concepts and system architectures examined.

Another risk is that some other person will complete largely the same research before I do. This has been addressed somewhat when this topic was selected. I asked Dr. Drexler if he was aware of any research in this area, and he said no [Drexler, 1995c]. The first part of the mitigation plan is to start with a subject that (apparently) no one else is currently looking at. The second part is to try to complete the research as soon as possible, reducing the period of exposure.

The risk of losing my work will be mitigated by keeping remote copies of the files for backup.

For the purposes of this dissertation, MNT is most significantly a theory in applied theoretical science, which states that certain technical performance parameters are achievable, including some novel functions for technical systems (see 2.3). If a fundamental flaw is identified which knocks down the entire edifice, then one major leg of this research is lost, and the research collapses. The probability of this happening is very small. [Drexler, 1992a] generally used conservative assumptions and significant safety factors, and many key conclusions are reinforced with multiple arguments [Drexler, 1981], [Drexler, 1986]. Furthermore, the arguments have been out in the technical community for some time [Drexler, 1981], and no one has identified a fatal flaw [Drexler et. al, 1992]. The basic strategy is to live with this risk.

Further analysis, however, is likely to refine the estimates for technical performance parameters achievable with MNT. There is a risk that some of these values will change over the course of the dissertation. The primary mitigation plan will be to build executable versions of the performance models, thus making recalculation as easy as possible, and to monitor the literature to identify any important changes in MNT performance estimates.

5.0 Conclusion

If approved, this research will seek to confirm or refute major unsubstantiated claims in the literature, to establish minimum estimates on the maximum performance of the class of MNT-based systems on a range of major space functions, and to generate at least one novel space system architecture feasible under scientific law.

The approach will be to conduct a bottom-up synthesis of MNT from previously defined space system architectures through grand strategies for space, and to then conduct a top-down definition and analysis of additional grand strategies, operational concepts, and system architectures which appear promising.

Comments on this research plan, suggestions to refine the process, and suggestions for specific system architectures, operational concepts and grand strategies to consider in the research are invited.

It is hoped that the Guidance Committee will approve this research plan, allowing the focus to shift to conducting the research.

References

[Ahren, H., 1995] "Molecular Modelling Software Manufacturers Improve Functionality," The Scientist, pp. 18-19, March 20, 1995

[Alexander, C., 1964] Notes on the Synthesis of Form, Harvard University Press, 1964.

[Amon, D.M., "The Impact of Molecular Engineering on Spacecraft Information Systems," in [Santoli, 1994a], pp. 303-310.

[Amato, I., 1991a] "A U.S. Lab Opens Doors To the Nanoworld," Science, Vol. 254, pp. 1302-1303, November 1991.

[Andrews, D.G., 1993] "Cost Considerations for Interstellar Missions," Acta Astronautica, Vol. 34, pp. 357-364, 1994

[Bennet, J., 1995] Personal communications, 6 July 1995.

[Binnig, G. and H. Rohrer, 1985] "Scanning Tunnelling Microscopy," Physica 127B, pp. 37-45, 1985.

[Borowski, 1990] "Nuclear Thermal Rocket Workshop Reference System -Rover/Nerva-," in [NASA Lewis Research Center, 1990], pp. 77- .

[Coppinger, R. J. and T. T. Toth-Fejel, 1995] "Fire and Ice on the Red Planet: Terraforming Mars with Drexlerian Assemblers," April 10, 1995, Draft.

[Crandal, B.C. (ed.), 1995] Nanotechnology: Speculations on the Emerging Culture of Abundance, MIT Press, 1995 (forthcoming).

[Crandall, B.C. and J. Lewis (eds), 1992] Nanotechnology: Research and Perspectives, MIT Press, 1992.

[Goure, D.,1994] "The Revolution that Isn't: RMA/MTR as Counter-Revolution" CSIS Paper, Center for Strategic and International Studies, Washington, D.C., November 2, 1994.

[Drexler, K.E., 1981] "Molecular engineering: an approach to the development of general capabilities for molecular manipulation," Proceedings of the National Academy of Science, Vol. 78, No. 9, pp 5275-5278.

[Drexler, K.E., 1986a] Engines of Creation: The Coming Era of Nanotechnology, Doubleday, New York, 1986.

[Drexler, K.E., 1986b] "Molecular Engineering, Assemblers and Future Space Hardware," American Astronautics Society, AAS-86-415.

[Drexler, K.E., 1988a] "Nanotechnology and the Challenge of Space Development," Proceedings of the 1988 National Space Society Conference.

[Drexler, K.E., 1988b] "Rod Logic and Thermal Noise in the Mechanical Nanocomputer," Proceedings of the Third International Symposium on Molecular Electronic Devices, Elsevier Science Publishers B.V., 1988.

[Drexler, K.E., 1991] "Engineering Parameters for materials and subsystems of aerospace interest produced by a mature molecular manufacturing technology," Prepared for the National Space Society annual conference, San Antonio, Texas, May 1991.

[Drexler, K.E., 1992a] Nanosystems: Molecular Machinery, Manufacturing, and Computation, Wiley Interscience, 1992.

[Drexler, K.E., 1992b] "Molecular Manufacturing for Space Systems: An Overview," in [Santoli, 1992a], pp. 401-405, October 1992.

[Drexler, K.E., 1992c] "Strategies for Molecular Systems Engineering," in [Crandall and Lewis, 1992] pp. 115-146.

[Drexler, K.E., 1995a] "Foreward" and "Introduction to Nanotechnology," in [Krummenacker and Lewis, 1995], pp. xvii-22.

[Drexler, K.E., 1995b] "Molecular Manufacturing as a Path to Space," in [Krummenacker and Lewis, 1995], pp. 197-205.

[Drexler, K.E., 1995c] personal communications.

[Drexler, K.E., C. Peterson and G. Pergamit, 1991] Unbounding the Future: The Nanotechnology Revolution, William Morrow, 1991.

[Drexler (et. al.), K.E., J.S. Foster, T. Handel, R.C. Merkle, M.D. Ward, 1992] "What Major Problems Need to Be Overcome to Design and Build Molecular Systems?," in [Crandall and Lewis, 1992], pp. 241-248.

[Eigler, D.M., and E.K. Schweizer, 1990] "Positioning single atoms with a scanning tunnelling microscope," Nature, Vol. 334, pp. 524-526.

[Feymann, R.P., 1960] "There's Plenty of Room at the Bottom," Engineering and Science, Vol. 23, pp. 22-36.

[Foresight Instutute, 1988] Foresight Update #3. 30 April 1988.

[Foresight Instutute, 1992] "Interview: Jim Bennett," Foresight Update #14, pp. 12-13, 15 July 1992.

[Freitas, R.A., Jr, 1980] "A Self-Reproducing Interstellar Probe," Journal of the British Interplanetary Society, Vol. 33, pp. 4251-264, 1980.

[Freitas, R.A., Jr. and W.P. Gilbreath, 1983] Advanced Automation for Space Missions, NASA Conference Publication 2255, National Technical Information Service, U.S. Department of Commerce, Springfield, VA, N83-15348.

[Friedman, G.J., 1995a] "An Expanded Agenda for SSI," SSI Update, Vol. XXI, Issue 1, pp. 1-4, 1995

[Friedman, G.J., 1995b] "Taxonomy of Non-Military, Space-Relevant Grand Strategies," Unpublished notes, 15 November 1995.

[Galushkin, A.I., P.N. Luskinovich, S.S. Nesmeyanov, V.I. Nikishin and V.D. Frolov, 1994] in [Santoli, 1994a], pp. 331-333.

[Goddard, R.H., with E.C. Goddard and G.E. Pendray (eds.), 1970] The Papers of Robert H. Goddard: Volume I: 1898 - 1924, McGraw-Hill, 1970.

[Godzik, A., 1995] "In Search of the Ideal Protein Sequence," Protein Engineering, Vol. 8 No. 5, pp. 409-416, 1995.

[Gray, M., 1992] Angle of Attack: Harrison Storms and the Race to the Moon, W.W. Norton and Company, 1992.

[Hall, J.S., 1993] "Overview of Nanotechnology," Internet Frequently Asked Questions (FAQ) on sci.nanotech Usenet newsgroup. (Adapted from papers by R.C. Merkle and K.E. Drexler), 1993.

[Hanson, P.A., 1992] "To the Stars with the Cytoskeleton?," in [Santoli, 1992a], pp. 415-420.

[Houbrechts, A., B. Moreau, R. Abagyan, V. Mainfroid, G. Préaux, A. Lamproye, A. Poncin, E. Goormaghtigh, J.M. Ruysschaert, J.A. Martial and K. Goraj, 1995] "Second-Generation Octarellins: To New De Novo ([[beta]]/[[alpha]])8 Polypeptides designed for Investigating the Influence of [[beta]]-Residue Packing on the [[beta]]/[[alpha]]-Barrel Structure Stability," Protein Engineering, Vol. 8 No. 3, pp. 249-259, 1995.

[Huang, D., H. Uchida and M. Aono, 1994] "Deposition and Subsequent Removal of Single Si Atoms on the Si(111)-7x7 Surface by a Scanning Tunneling Microscope," Journal of Vacuum Science and Technology B, vol. 12 no. 4, pp. 2429-2433, July/August 1994.

[Instutute for Molecular Manufacturing (IMM), 1991] The NanoReporter, Vol.1 No.1, 21 October 1991.

[Janson, S.W., H. Helvajian and E.Y. Robinson 1993] "The Concept of "Nanosatellite" for Revolutionary Low-Cost Space Systems," IAF-93-U.5.573, from the 44th Congress of the International Astronautics Federation, pp. 5, 1993

[Janson, S.W., 1994] "Chemical and Electrical Micropropulsion Concepts for Nanosatellites," AIAA-94-2998, pp.14, 1994

[Klir, G.J., 1985] Architectures of Systems Problem Solving, Plenum, 1985.

[Kritsky, M.S. and N.P. L'vov, 1992] "Flavoproteins as Natural Prototypes of Molecular Electronic Devices with Photocontrolled Conductivity," in [Santoli, 1992a], pp. 421-426.

[Krummenacker, M. and J. Lewis (eds.), 1995] Prospects in Nanotechnology: Toward Molecular Manufacturing, John Wiley & Sons, 1995.

[Landman, H.A., 1995] "Paths to Nanotechnology," in [Krummenacker and Lewis, 1995], pp. 113-127.

[Luskinovich, P., 1994] "Stages of Technical Evolution," in [Santoli, 1994a], pp. 343-344.

[Ludewig, 1990] "Liquid Annulus," in [NASA Lewis Research Center, 1990], Proceeding of the Nuclear Thermal Propulsion Workshop, sponsored by NASA Lewis Research Center, July 10-12, 1990, NASA Conference Publication 10079, p.286 - .

[Mackenzie, B.A., 1995] "Bootstrapping Space Construction with Micro Rovers and High Tensile Boot Laces (Tethers)," Proceedings of Space Manufacturing 10, Barbara Faughnan, ed., American Institute of Aeronautics and Astronautics, forthcoming 1995.

[Mallon, J., 1992] "Nanotechnology from a Micromachinist's Point of View," in [Crandall and Lewis, 1992] pp. 215-240.

[Mallove, E.F., F.L. Forward and J. Paprotny, 1980] "Interstellar Travel and Communication: A Bibliography," Journal of the British Interplanetary Society, Vol. 33, pp. 201-248, 1980.

[Maryniak, G., undated] "Self Replicating Machines for Space," Space Studies Institute, undated (latest included citation 1983).

[McKendree, T. L., 1988] "Heuristics for Playing Board Games," assignment in AE/ISE 599a at USC, unpublished, 1988.

[McKendree, T. L., 1989a] "The STM as an Architected System," assignment in AE/ISE 599b at USC, unpublished, February 1989.

[McKendree, T. L., 1989b] "Disneyland as an Architected System," assignment in AE/ISE 599b at USC, unpublished, March 1989.

[McKendree, T. L., 1989c] "Operation Overlord as an Architected System," assignment in AE/ISE 599b at USC, March 1989, in [Rechtin, 1989].

[McKendree, T. L., 1989d] "Delta as an Architected System," assignment in AE/ISE 599b at USC, unpublished, March 1989.

[McKendree, T. L., 1993] "Educating Systems Engineers for an Impending Metatechnology in Product and Process Systems," Proceedings of the Third Annual International Symposium of the National Council on Systems Engineering, James E. McAuley and William H. McCumber, eds., National Council on Systems Engineering, forthcoming 1993.

[McKendree, T. L., 1995] "Planning Scenarios for Space Development," Proceedings of Space Manufacturing 10, Barbara Faughnan, ed., American Institute of Aeronautics and Astronautics, forthcoming 1995.

[Merkle, R.C., 1990] "NASA and Self-Replicating Systems," Foresight Update #9, 1990.

[Merkle, R.C., 1992] "Self Replicating Systems and Molecular Manufacturing," in [Santoli, 1992a], pp. 407-413, October 1992.

[Merkle, R.C., 1995a] "Design-Ahead for Nanotechnology," in [Krummenacker and Lewis, 1995], pp. 23-52.

[Merkle, R.C., 1995b] "How long will it take to develop nanotechnology?," http://nano.xerox.com/nanotech/howlong.html

[Moravec, H.P., 1977] "A Non-Synchronous Orbital Skyhook," Journal of the Astronautical Sciences 25, October/December 1977.

[Morgan, C.R., 1994] "Terraforming With Nanotechnology, in [Santoli, 1994a], pp. 311-318.

[Musgrove, C., J. Perry, R.C. Merkle and W.A. Goddard III, 1992] "Theoretical analysis of a site-specific hydrogen abstraction tool," Nanotechnology, April 1992.

[Nadler, G., 1981] The Planning and Design Approach, Wiley Interscience, 1981.

[Nadler and Hibino, 1990] Breakthrough Thinking: Why We Must Change The Way We Solve Problems, And The Seven Principles To Achieve This, Prima Publising & Communications, 1990.

[NASA Lewis Research Center, 1990] Proceeding of the Nuclear Thermal Propulsion Workshop, sponsored by NASA Lewis Research Center, July 10-12, 1990, NASA Conference Publication 10079.

[Nelson, M. and C. Shipbaugh, 1995] The Potential of Nanotechnology for Molecular Manufacturing, RAND Manuscript, MR-615-RC, RAND, 1995.

[Nussinov, M.D., S.V. Lysenko and V.V. Patrikeev, 1994] "Terraforming of Mars Through Terrestrial Microorganisims and Nanotechnological Devices," in [Santoli, 1994a], pp. 319-320.

[Nussinov, M.D. V.I. Maron and S. Santoli, 1994] "The Evolution of Soliton-Like Impulse Paradigm of Matter: Perspectives on Unity in the Nanotechnology/Nanobiology Dichotomy," in [Santoli, 1994a], pp. 321-324.

[O'Neill, G.K., 1974a] "A Lagrangian community?," Nature, p. 636, August 23 1974.

[O'Neill, G.K., 1974b] "The Colonization of Space," Physics Today, pp. 32-40, September 1974.

[O'Neill, G.K., 1975] "Space Colonies and Energy Supply to the Earth," Science, pp. 943-947, 5 December 1975.

[O'Neill, G.K., 1976] The High Frontier, Space Studies Institute Press, 1989.

[Pergamit, G., 1995] "Senior Associates Confer at Foresight," Foresight Update #20, pp. 11,14, 1 February 1995.

[Peterson, C.L., 1992] "Nanotechnology: Evolution of the Concept," in [Santoli, 1992a], pp. 395-400, October 1992.

[Pinneo, M., 1995] "Diamond Growth: Today and Tomorrow," in [Krummenacker and Lewis, 1995], pp. 147-172.

[Raymond, A., 1951] The Well Tempered Aircraft, London: Royal Aeronautical Society, 1951.

[Rechtin, E. (Ed.), 1989] "A Collection of Papers on Systems Architecting," Unpublished graduate report, University of Southern California, Los Angeles.

[Rechtin, E., 1991] Systems Architecting, Creating and Building Complex Systems, Prentice Hall, 1991.

[Rechtin, E., 1994] "Foundations of Systems Architecting," Systems Engineering, Vol. 1 No. 1, pp. 35-42, July/September 1994.

[Regan, L. and W. DeGrado, 1988] "Characterization of a helical protein designed from first principles," Science, 241, pp. 976-978.

[Regis, E., 1995] Nano! Remaking the World Atom by Atom, Little, Brown and Company, 1995.

[Rowe, W.D., 1965] "Why Systems Science and Cybernetics?," IEEE Transactions on Systems Science and Cybernetics, Vol. 1 No 1., pp. 2-3, November 1965.

[Sackheim, R.L., R.S. Wolf and S. Zafran, 1991] "Chapter 17: Space Propulsion Systems," in [Wertz and Larson, 1991], pp. 579-606.

[Sander, C., G. Vriend, F. Bazan, A. Horovitz, H. Nakamura, L. Ribas, A.V. Finkelstein, A. Lockhart, R. Merkl, L.J. Perry, S.C. Emery, C. Gaboriaud, C. Marks, J. Moult, C. Verlinde, M. Eberhard, A. Elofsson, T.J.P. Hubbard, L. Regan, J. Banks, R. Jappelli, A.M. Lesk and A. Tramontano] "Protein Design on Computers. Five New Proteins: Shpilka, Grendel, Fingerclasp, Leather, and Aida," Proteins: Structure, Function, and Genetics, Vol 12, pp. 105-110, 1992.

[Santoli, S. (ed.), 1992a] Journal of the British Interplanetary Society Vol. 45 No. 10: Nanotechnology, 1992.

[Santoli, S., 1992b] "Enroute to Basic Machine Members for Molecular Manufacturing: Hypotheses for Rigid-Link Nanorobotic Arms," in [Santoli, 1992a], pp. 427-432.

[Santoli, S. (ed.), 1994a] Journal of the British Interplanetary Society Vol. 47 No. 8: Nanotechnology (Part 2), 1994.

[Santoli, S., 1994b] "System-Level Physics of Autonomous Nanorobots for Hard Chemistry and Wave Packet Engineering," in [Santoli, 1994a], pp. 335-340.

[Santoli, S., 1994c] "Developing the Concept of a Rigid-Link Nanorobotic Arm Working in a Dense Fluid," in [Santoli, 1994a], pp. 341-342.

[Schwartz, P., 1991] The Art of the Long View: Planning for the Future in an Uncertain World, Currency/Doubleday, 1991.

[Sheka, E., V. Khavryuchenko, P. Luskinovich, V. Nikishin, E. Nikitina, V. Ogenko, I. Ryzhikov, and I. Gorelkin, 1994] "Computer Modelling in an Electric Field: Nanometric Design with Scanning Tunnelling Microscope," in [Santoli, 1994a], pp. 325-330.

[Simon, H.A., 1981] Sciences of the Artificial, The MIT Press, 1981.

[Studt, T., 1995,] "Molecular Modeling Software Changes Research Techniques," R & D Magazine, pp. 77-79, February 1995.

[Taniguchi, N., 1983] "Current Status in and Future Trends of Ultraprecision Machining and Ultrafine Materials Processing," Annals of CIRP, Vol. 32 No. 2, 1983.

[Toth-Fejel, T. T., 1984] "Self-Test: Murphy's Law Versus Electronic Automata, >From Simple Circuits To Self-Reproducing Robotic Systems," Master of Science Thesis in Electrical Engineering, Notre Dame, June 1984.

[Wertz, J.R. and W.J. Larson, 1991] Space Mission Analysis and Design, Space Technical Library (Kluwer Academic Publishers), 1991.