27 February 2020

Research pick: Slimy computation - "Slime mould foraging: an inspiration for algorithmic design"

Nature has provided a great deal of inspiration for computer scientists developing search algorithms and ways to solve complicated problems with as little computing power as possible. Ant colonies, beehives, bat hunting, and now slime mould foraging can be used as models on which an algorithm can be constructed.

Writing in the International Journal of Innovative Computing and Applications, Anthony Brabazon and Sean McGarraghy of the University College Dublin, Dublin, Ireland, explain how 99.5% of the living things on earth lack neurones and yet are proven success stories despite what we, as neuronal creatures, might whimsically perceive as a deficiency. One group of organisms that have been rather successful for millions of years are the so-called slime moulds. The term is an informal name for several different groups of organisms that are actually unrelated. They are not moulds, rather they are organisms that can live freely as single cells, but under certain conditions will form communicating aggregates that work in concert as if they are a multicellular reproductive structure.

The team explains that the plasmodial slime mould Physarum polycephalum, which forms from aggregates of individual amoebae, encases itself in a thin membrane and can act as a single organism. The researchers explain how “Inspiration has been drawn from some of its foraging behaviour to develop algorithms for graph optimisation.” They report examples of the algorithms that can be developed and make suggestions as to how future research might proceed to utilise the benefits and minimise any limitations.

Of course, the slime mould itself is, despites its lack of neurons, carrying out computations all the while, chemical computations, you might say. So, in a sense modelling its behaviour in an algorithm is an excellent foundation.

“Of course,” the team concedes, “it is also important to note that the developed algorithms are very simplified representations of (the imperfectly understood) real-world foraging behaviours of P. polycephalum and other slime moulds and doubtless future biological research concerning these organisms will open up new avenues of investigation.”

Brabazon, A. and McGarraghy, S. (2020) ‘Slime mould foraging: an inspiration for algorithmic design’, Int. J. Innovative Computing and Applications, Vol. 11, No. 1, pp.30–45.

No comments: