Article Preview
TopIntroduction
The plasmodium stage of the giant single-celled true slime mould Physarum polycephalum is demonstrated by a complex morphological interaction with its environment. The growth and migration of the amorphous acellular organism is affected by the presence of both attractants and repellents. The plasmodium forms a spatially efficient protoplasmic tube transport network, within which nutrients are transported within the mass of the organism. The efficiency of this network is a trade-off in terms of material cost, path redundancy and resilience to damage. The topography of the transport network may be viewed as a living ‘circuit' pattern by which is constructed in a bottom-up manner in a system comprising relatively simple components, guided by local stimuli. This contrasts with classical top-down methods of circuit design which utilise globally applied templates to enforce connectivity patterns between components.
Growing Physarum circuits can be controlled by various means (Adamatzky, 2010b), including physical modification of growth substrate (Figure 1a), chemo-repellents (Adamatzky, 2010c) (Figure 1b) or even high-concentration of chemo-attractants (Figure 1c). When oat flakes colonised by plasmodium are placed in the northern part of a 12x12cm Petri dish filled with agar, Physarum cells are attracted by virgin oat flakes in the southern part of the dish. Physically removed substrate, sodium chloride, and sugar in high concentration all act as repellents. Also, as non-contact methods of geometrically shaped illumination (Adamatzky, 2009) (Figure 2). In this case, white light acts as repellents and 8-shaped window is cut in the aluminium foil underlying the Petri dish. The 8-shaped window is illuminated by Seiko El Sheet (see details of the experiments in Adamatzky, 2009). Using of the techniques modifying physical structure of substrate or its chemical components would make slime mould circuits non-reusable, because chemical species continue to diffuse in the substrate. Moreover, even the case of simple removal of substrate would not be suitable because they do not allow for reconfiguration of the circuits. Non-invasive control techniques, as, e.g., with light (Adamatzky, 2009) would be appropriate; however they are rather unreliable and do not always predictable control of the growing slime mould.
Figure 1. Controlling Physarum by “conventional'' means: (a) Physical removal of substrate: 8-shaped piece of agar plate removed is visible as black shape, (b) Sodium chloride: position of two grains of salt is shown by stars, and (c) High concentration of sugar: several grains of sugar are placed in the positions indicated by stars. (Images of Petri dishes are scanned on Epson Perfection 4490. Images are taken 24h after inoculation).
Figure 2. Controlling Physarum by “conventional'' means. Oat flakes colonised by plasmodium and virgin (oat flakes are positioned similarly as in Figure 1): (a) Photograph of the experimental Petri dish (under-dish illumination is switched off) is done with FujiFilm Fine-Pix digital camera, and (b) Scheme of the experiment: seeds of Physarum are shown by rhombs, destination oaf flakes by discs, major protoplasmic tubes reflecting trajectories of propagating acting zones by lines.
In this paper we investigate the possible exploitation of biologically derived circuit pattern construction by interfacing the plasmodium stage of Physarum polycephalum with the behaviour of a spatially implemented general purpose analog computer, the EAC.