Search

This paper discusses a novel approach to managing complexity in a large self-assembled system, by employing the self-assembling components themselves to address the complexity. Challenges are discussed, including the question of how to deal with elements that are assembled in different orientations from each other, and a solution based on the idea of introspective circuits is described. A methodology for determining a single cell’s orientation from an adjacent cell is given. An algorithm is then described for using such re-oriented edge cells to determine orientation of more-interior cells, thus allowing re-orientation of an entire 2D region of cells. Test procedures are described, and results are presented to show better-than-linear time performance (O(sqrt(n))). The significance of this work is discussed, not only in terms of managing arrays of dis-oriented cells, but more importantly as an example of the usefulness of local, distributed self-configuration to create and use introspective circuitry. Finally, future work is discussed, including extension to 3D collections of cells.