As a plant's root bursts from its seed, it has no knowledge of the distribution of nutrients in the soil below itself, and yet, it must grow in such a way as to secure access to the numerous nutrients on which the plant relies for its survival. Not possessing a centralized decision making organ such as a brain, the root relies instead on distributed sensing and acting by its apexes, the cell agglomerates at the tips of every root segment. Seeking to replicate the mechanisms by which roots achieve their remarkable success in exploring the soil, we created a simulator of soil and root dynamics. Apex behaviours were then automatically designed on it through an optimization process: a swarm of homogeneous neural networks, acting in parallel and communicating through both direct and stigmergic means, was made to succeed in producing global structures of maximal benefit to the whole plant. In this presentation I will provide an overview of the work conducted in this project, and its main results, namely in the transfer of apex behaviours into a controller for deployment of sensor webs on a planetary surface.
plant roots, swarm intelligence
Luis F. Simoes, Cristina Cruz, Rita A. Ribeiro, Luis Correia, Tobias Seidl, Christos Ampatzis, Dario Izzo. (2011)
Path Planning Strategies Inspired by Swarm Behaviour of Plant Root Apexes. Technical Report 09/6401 of European Space Agency, Advanced Concepts Team.
Ariadna Final Report.