Computing with Biological Dynamics
This volume of the International Journal of Unconventional Computing is dedicated for The International Workshop on Computing with Spatio-Temporal Dynamics 2010 (CSD10), a satellite workshop of the international conference on Unconventional Computation 2010 (UC10), in The University of Tokyo, Hongo Campus, Japan, on June 21–25, 2010. The workshop aimed to discuss the information processing capabilities in nature at all scales (physical, chemical, and biological) and how we could exploit them by bringing together “explorers“ who are struggling to find those “hidden gems” of natural computing.
The Steering Committee of the series of workshops Physics and Computation includes Caslav Brukner (University of Vienna), Cristian S. Calude, (University of Auckland), Gregory Chaitin (IBM Thomas J.Watson Research Center), José Félix Costa (Technical University of Lisbon) and István Németi (Hungarian Academy of Sciences).
Since the first proposal of biological computers , it has been now widely accepted that biological systems exhibit intriguing computing capabilities, which are difficult to duplicate on digital computing systems. Considering that living systems have no logic gate as in conventional computers, what kind of mechanisms support biological intelligent behaviour? How those biological computations can be mimicked on computers? Is there any way that natural systems can be harnessed for man-made computing? These questions have been guiding engineers of unconventional computing systems. Although we have now obtained many technologies which may help us achieve these dreams, only way to tackle them is just simply to construct systems — on the computer screen as well as under the microscope.
Authors in this special issue also are those who are challenging the big questions. Especially this special issue features papers from “soft”, “hard”, and “wet” unconventional computers in the following pages.
First, “soft” unconventional computer, i.e. a software-based unconventional computing model, is presented by Jeff Jones. His model particularly focuses on a kind of amoeboid slime mould, Physarum polycephalum. This organism, despite its single cellular structure, is known to show smart behaviour, such as maze-solving , road planning [1, 8], and cellular memory . So far several biological computing systems have been built using the organism [2, 9]. It is often modeled as multi-agent systems in which simplistic agents are continuously interacting with other agents. A global behaviour then emerges through interactions between agents. Slime mould’s behaviour can be described similarly—Local parts of a slime mould cell showing oscillatory behaviour are interacting each other through proptoplasmic streaming. Any change in a local part (external stimulation, such as light and temperature) propagates through the protoplasmic streaming to other parts, and then results in a change at a whole-cell level (e.g escape from light, migration towards warmer area). Jones proposes a “particle swarm” model that approximates various kinds of behaviour of slime mould. Each particle incorporates a simple sensor to detect virtual chemical gradient like pheromone deposited by other particles. A swarm of particle react to the chemical gradient as well as external stimuli. This well approximates known behaviour of slime mould, and it may be speculated to be useful for the control smart materials, such as mobile robots.
A “hard” unconventional computer is then investigated by Takuya Umedachi and his colleagues. They have designed and constructed coupled mobile robots inspired by Physarum slime mould. The Physarum-inspired robot consists of a ring of oscillating components with a air-filled balloon in the centre. The balloon interconnects components and therefore the whole robot shows coordinated locomotion such as migration towards attractant. This corresponds to the protoplasm of the Physarum plasmodium allows local parts of a slime mould to interact with other parts. If the balloon is burst, the oscillating units lose interactions and therefore they show no locomotion, just like the slime mould lose synchronisation between local parts when protoplasmic streaming is stopped .
The third aspect, “wet” unconventional computing is first explored by Masashi Aono and his colleagues. They employed the Physarum slime mould as a computing component of artificial neural network. Following to their previous studies , they have investigated the scalability of Physarum-based computing systems for solving 8-city travelling salesman problem (TSP). The slime mould is not the only organisms which can be used to implement neurocomputers. Kazunari Ozasa and his colleagues drew focus to a kind of plankton, Euglena gracilis. This organism exhibits phototactic reaction to external light stimuli. A population of Euglena is placed in a microstructure and interfaced to Hopfield-Tank neural network via optical feedback system. The blue light exposure to the microstructure triggers changes of Euglena swimming patterns, which will be sent to the neural network on a computer as feedback. The interactions between them lead to finding of best solution of TSP as well as searching of multiple solutions. These unconventional computers undoubtedly exemplify the possibilities clear examples of hybrid systems that exploit adaptive information capabilities of living systems.
Computing models presented in this special issue are, however, one aspect of unconventional spatio-temporal computing systems discussed at CSD10. How can non-living chemical systems be adapted for computing applications? This aspect is discussed in a sister special issue of CSD10 in the International Journal of Nanotechnology and Molecular Computation (Volume 3, Issue 1). Interested readers are highly recommended to refer to the special issue as well.
 A. Adamatzky and J. Jones. (2010). Road planning with slime mould: If physarum built motorways it would route m6/m74 through Newcastle. International Journal of Bifurcation and Chaos, 20(10):3065–3084.
 M. Aono and M. Hara. (2008). Spontaneous deadlock breaking on amoeba-based neurocomputer. Bio Systems, 91(1):83–93.
 M. Aono, Y. Hirata, M. Hara, and K. Aihara. (2009). Amoeba-based chaotic neurocomputing: Combinatorial optimization by coupled biological oscillators. New Generation Computing, 27(2):129–157.
 M. Conrad. (1992). Molecular computing: the lock-key paradigm. Computer, 25(11):11– 20.
 Y. Miyake, M. Yano, and H. Shimizu. (1991). Relationship between endoplasmic and ectoplasmic oscillations during chemotaxis of physarum polycephalum. Protoplasma, 162(2):175–181.
 T. Nakagaki and H. Yamada. (2000). Maze-solving by an amoeboid organism. Nature, 407(28 September):470.
 T. Saigusa, A. Tero, T. Nakagaki, and Y. Kuramoto. (2008). Amoebae anticipate periodic events. Physical Review Letters, 100(1):018101.
 A. Tero, S. Takagi, T. Saigusa, K. Ito, D. P. Bebber,M. D Fricker, K.Yumiki, R. Kobayashi, and T. Nakagaki. (2010). Rules for biologically inspired adaptive network design. Science, 327(5964):439–442.
 S. Tsuda, K.-P. Zauner, and Y.-P. Gunji. (Feb 2007). Robot control with biological cells. Bio Systems, 87 (2-3):215–23.