AI Seminar: Kazuhiro Saito

[Eric D. Klavins]

AI Seminar
----------
Tuesday Feb. 15, 2000
4:00 pm - 5:00 pm
175 ATL (Large Conference Room)

Conformational Switching in Self-Assembling Mechanical Systems
--------------------------------------------------------------
Kazuhiro Saitou
Assistant Professor
Department of Mechanical Engineering and Applied Mechanics
University of Michigan

Abstract:

A study of one-dimensional self-assembly of a type of mechanical
conformational switches, minus devices, is presented where assembly occurs
via the sequential mating of a random pair of parts selected from a part
bin, referred to as sequential random bin-picking. Parametric design
optimization of the minus devices via a genetic algorithm maximizing the
yield of a desired assembly, and rate equation analyses of the resulting
designs, reveal that the minus devices facilitate the robust yield of a
desired assembly against the variation in the initial fraction of the part
types, by specifying a fixed assembly sequence during the self-assembling
process. It is also found that while the minus devices can ``encode''
some assembly sequences, encoding other assembly sequences requires the
use of another type of conformational switches, plus devices. To
investigate the ``encoding power'' of these conformational switches, a
formal model of self-assembling systems, one-dimensional self-assembling
automaton, is introduced where assembly instructions are written as local
rules that specifies conformational changes realized by the conformational
switches. It is proven that the local rules corresponding to the minus
and plus devices, and three conformations per each component, can encode
any assembly sequences of a one-dimensional assembly of distinct
components with arbitrary length.

Keywords: mechanical conformational switches, self-assembling mechanical
systems, genetic algorithms, assembly sequences, assembly grammars.