Kinematic Self-Replicating Machines
© 2004 Robert A. Freitas Jr. and Ralph C. Merkle. All Rights Reserved.
Robert A. Freitas Jr., Ralph C. Merkle, Kinematic Self-Replicating Machines, Landes Bioscience, Georgetown, TX, 2004.
6.4.5 Embody Principles of Good Design
Any proposed design should embody the many principles of good engineering design, of which there is space here to briefly mention only a few:
(1) Assumptions and Design Tradeoffs Explicitly Stated. An important principle of good design is to make all assumptions and design tradeoffs explicit. This helps readers calibrate the proposed design against their own understanding of the technological possibilities, illuminates areas most needing further research, and helps others who may wish to adopt different sets of tradeoffs in hopes of exploring other equally promising volumes of the molecular assembler design space.
(2) Seek Broad “Attractors” in the Design Space. Wherever possible, design solutions should be chosen from sets of possible design choices which occupy the broadest possible volume of design space. Such solutions therefore coexist with many closely related viable design alternatives, allowing both flexibility and robustness in the final design. The demonstrated existence of adequate “design space” (Section 5.1.9) is the single most important reason for believing that feasible designs for molecular assemblers exist. Also important is the concept of modular design and the feasibility of multiple alternative implementations of a given module, with the avoidance of overly “clever” designs that try to solve too many problems at once.
(3) Exemplify Backward Chaining in Design. Many existing designs for molecular assemblers are more complex and appear more difficult to manufacture than the Merkle-Freitas design (Section 4.11.3) – a design which is simpler while still exemplifying the principle of backward chaining (Section 6.2.1) in design. A simple design such as this should help pave the way toward future designs which will be simpler still, and even more amenable to manufacture, once further efforts at additional backward chaining have been made.
For example, the Merkle-Freitas design was chosen for its likely intermediate position in the design and development timeline for molecular assemblers. It appears to be one of the least complicated assemblers that could conveniently be specified to molecular detail and yet possesses sufficient capability to allow for the design and manufacture of progressively more complex and capable assembler systems, leading eventually to a molecular manufacturing system of very generalized capability. On the other hand, the same system could be sufficiently precisely specified that it would also permit convenient backward chaining to simpler pre-assembler system designs, a process that is hoped would eventually merge with our ever-expanding contemporary nanoscale technology base, finally providing a clear and complete pathway for the development for molecular manufacturing systems.
(4) Buildability of the Design. Buildability is an explicit design objective. In the broadest sense, a design may be said to be buildable if, given the proper manufacturing tools and techniques, it could be expressed in hardware. A design which is buildable in this broadest sense must violate no known laws of physics or chemistry and should reflect customary principles of sound engineering design. In a more restrictive sense, a design might also be considered feasible to build if, given certain foreseeable extensions of current technology in directions that can readily be envisioned today, it could be expressed in hardware using those foreseeable technological extensions. In the narrowest sense, a design might be considered buildable only if it can be built in the laboratory or factory today, using nothing but currently available techniques and existing technology and equipment with no further improvements. Buildability in this most narrow sense should not be an objective of a focused design effort intended to demonstrate feasibility of a molecular assembler. Rather, the principal objective should be to create a design that is buildable in the more restrictive sense. This design can then be used, first, to guide the development of successor designs which more closely approach today’s building techniques, and second, to simultaneously guide the development of new building techniques that can more closely approach the designs that we believe would be useful to build, until a sufficient overlap arises between these two sets that it becomes possible to build, as physical hardware, the first working molecular assembler.
(5) Specificity of the Design. A good system design should describe all important components, systems, and operations with enough specificity to permit rigorous analysis and evaluation by competent engineers. For example, Merkle  has described the wide range of details which should be present in any serious molecular assembler design proposal. These details include: (1) the type and construction of the computer, (2) the type and construction of the positional device, (3) the set of chemical reactions that take place at the tip, (4) how compounds are transported to and from the tip, (5) how the compounds are modified (if at all) before reaching the tip, (6) the class of structures that can be built, (7) the environment in which the device may operate, (8) the method of providing power, (9) the requirement (or not) for a barrier that partitions internal and external environments, (10) the type of barrier that can prevent unwanted changes in the internal environment in the face of changes in the external environment, (11) the nature of the external and internal environments, (12) the transport mechanisms that move material across the barrier, (13) the transport mechanism used in the external environment, and (14) a mechanism that allows the assembler to receive broadcast instructions. (Notes Merkle : “The presence of a receiver is not mandatory for the manufacture of interesting structures (as demonstrated by a fertilized egg), but it is extremely useful if we are considering a general purpose device able to make a wide variety of different products. Not only does a receiver allow us to re-program the assembler for specific tasks, it also allows us to significantly simplify the computational element.”)
Last updated on 1 August 2005