© 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.

4.11.3.3 The Broadcast Architecture for Control

Von Neumann’s proposed architectures for self-replication [3], the replicating lunar factory designs proposed for NASA [2] in 1980 by Freitas and by von Tiesenhausen and Darbro, Drexler’s original assembler proposal in 1986 [199, 208], and living systems all carry a complete set of plans for the system onboard the replicating system, in some form of internal memory. This is not a logical necessity in general manufacturing systems, as was observed by the 1980 NASA study (Freitas and Gilbreath [2], p. 215): “It is quite possible to imagine the lunar factory operating nonautonomously...the in situ computer used simply as a teleoperation-management system for operations controlled directly by Earth-based workers [with] information necessary to accomplish [replication] supplied from outside. An alternative would permit the on-site computer to handle mundane tasks and normal functions with humans retaining a higher-level supervisory role.”

Broadcast instructions (teleoperation) can enable system replication without internally stored instructions (Section 2.3.6), an arrangement which has been termed the “broadcast architecture” [208-210], first described in the parallel-operation nanotechnological context in Section 16.3.2(a) of Nanosystems [208]: “Broadcast instructions. Replacing computers and stored instructions with broadcast instructions can simplify molecular manufacturing systems. Signals can be broadcast to molecular mechanical systems in several ways, but signaling by modulated pressure via mechanical transducers is both simple and adequate.” Similar conclusions have been reached by space technologists who suggest using remote-control procedures (e.g., “teleoperation,” “telepresence,” “telerobotics,” or “teleautonomy” [18]; see also Section 3.9) to reduce the high cost of items that would otherwise have to be expensively raised into orbit.

If we separate the “constructor” from the “computer” and allow many individual constructors to receive broadcast instructions from a single central computer, then each constructor need not remember the plans for what it is going to construct [209]. Rather, it can simply be told what to do as it does it (Figure 4.44). This approach eliminates the requirement for an integral repository of plans within the constructor (which is now the only component that must replicate) and also eliminates almost all of the mechanisms involved in decoding and interpreting those plans. This general approach is similar to the SIMD architecture of the Connection Machine [2326], in which a single complex central processor broadcasts instructions to a large number of very simple processors. Storing the program, decoding instructions, and other common activities are the responsibility of the single central processor; while the large number of small processors need only interpret a small set of very simple instructions, thus allowing those small processors to retain comparatively simple structure and function. It is interesting that the living cell also employs the broadcast architecture [209] on a massively-parallel scale, using chemical rather than electrical or acoustic signaling: the DNA-containing nucleus acts as a central computer and data repository that broadcasts chemical instructions in the form of mRNA messenger molecules to up to ten million [228] structurally simpler manufacturing devices distributed throughout the eukaryotic cytoplasm, called ribosomes. Following the instructions received from the messenger molecules, the ribosomes manufacture all cytoplasmic proteins using unidimensional positional assembly (Section 4.2).

The specific advantages of the broadcast architecture are that: (1) it reduces the size and complexity of the self-replicating component, (2) it allows the self-replicating component to be rapidly redirected to build something novel, and (3) it is “inherently safe” (Section 5.11) because the individual constructors lack any capability to function autonomously, a feature that is particularly clear when the central computer is macroscopic and under our direct control. For these reasons we explicitly adopt the broadcast architecture for control in our proposed design for a molecular assembler.

Last updated on 1 August 2005