Locomotion as a Result of Displacement of Resources Payam Zahadat and Thomas Schmickl University of Graz, Austria payam.zahadat@uni-graz.at Nature has invented several different mechanisms to im- plement locomotion. In some organisms, locomotion is a result of coordination of muscle movements in distinct parts of the body, e.g. walking behavior. It has been a source of in- spiration for various artificial systems, e.g. Owaki and Ishig- uro (2017). In other natural organisms, locomotion is an emergent effect of local interactions between many simple agents moving as aggregated swarms in their environment, e.g. amoeboid movement in eukaryotic cells, amoeba, and slime-molds. In such continuous locomotions, flows of sub- strates form pseudopodium towards favorable regions of the environment. The substrates that form the pseudopodium are taken from other parts of the organism located in less favorable regions. This displacement of substrates leads to a crawling-like movement of the organism. A compar- atively similar mechanism is the locomotion of leaderless flocks of animals which have inspired several methods of self-organized locomotion for artificial swarms (Hornby and Pollack, 2001; Varughese et al., 2016). A related field is the formation of shapes in artificial swarms via reposition- ing of mobile agents, e.g. (Rubenstein et al., 2014). Here we report on a previous work (Zahadat and Schmickl, 2017) demonstrating a self-organized amoeboid-like locomotion controlled by a variation of a plant-inspired algorithm, called Vascular Morphogenesis Controller (VMC) (Zahadat et al., 2017). The VMC is initially introduced to guide the growth of artificial structures via a self-organized process, in the context of project flora robotica. The algorithm is inspired from competition of different branches in a plant enabling the plant to reach the areas with better conditions (e.g. more light). Here a variation of VMC is used to direct the loco- motion of an organism by the means of growing branches towards favorable regions while the non-favorable branches are retracted leading to an amoeboid-like locomotion. Vascular Morphogenesis Controller The morphology of a plant changes over time as a result of interplays between the encoded genome and the envi- Figure 1: a schematic example VMC-guided organism. ronment (Scarpella et al., 2010). The branches of a plant seek local resources such as light from their immediate ex- ternal environment and receive shares of limited common resources, i.e. water and minerals, from roots distributed by their internal vessel system. The common resource reaching a branch is used for the growth of the branch and produc- tion of new parts, i.e. new branches. The vascular system of a plant is dynamic and changes based on competitions for common resource between different branches (Leyser, 2011). A hormone called auxin is produced at the tip of each branch depending on its success in accessing local environ- mental resources (e.g. the amount of light). The branches in better locations produce larger amounts of the hormone. The hormone flows root-wards and regulates the quality of the vessels in its way. A branch with higher quality of ves- sels, receives larger shares of the common resource leading to further growth. This in turn may lead to a positive feed- back by positioning the branch in an even better location, more production of the hormone and better vessels, and thus even further growth of the branch. A high share of the lim- ited common resource for successful branches means lower shares of the resource for the others which maybe lead to decrease or stop of growth in the branches in less favorable regions of the environment. The VMC abstracts the above concepts into a decentral- ized algorithm acting on growing acyclic directed graphs (see Fig. 1). The success of a leaf i in a graph de-