Hey Physarum! Can you Perform SLAM?
Vicky S. Kalogeiton, Dim P. Papadopoulous and George Ch. Sirakoulis
In this paper, an innovative bio-inspired unconventional approach to tackle with the Simultaneous Localization and Mapping (SLAM) task is presented. The proposed method draws inspiration from the slime mold Physarum polycephalum by utilizing the computational tool of Cellular Automata (CA). In particular, a fully autonomous robot, equipped only with an omni-directional camera, explores and maps successfully an indoor unknown terrain by adopting the behaviour of the microorganism Physarum polycephalum, namely its olfaction, its propagation and foraging process. The Physarum’s sense of olfaction corresponds to the robots unique sensor, namely an omni-directional camera and the foraging as well as the movement procedure of the organism in terms of robots SLAM is matched with the detection and mapping of an unknown space and the movement strategy, respectively. Moreover, the diffusion field of the plasmodium corresponds to the robots field of view. In order to evaluate the proposed approach several experiments are drawn, that indicate the ability of the presented model to effectively and efficiently achieve its goals. The obtained results were compared to the corresponding ones produced by the random movement algorithm as well as by an exhaustive search algorithm. In all the examined cases, the presented simulation results reveal the strength and the superiority of the proposed method.
Keywords: Physarum polycephalum, simultaneous localization and mapping (SLAM), cellular automata, bioinspired, biomimetics, robot, camera, modeling