Local Sensory Feedback Generates Various Wave Gaits in Multi-legged Robots via Embodied Sensorimotor Interaction

According to the species and situations, multi-legged animals show three wave-like ipsilateral interlimb coordination. The swing leg movements propagate from posterior to anterior (direct-wave), from anterior to posterior (retrograde-wave), and in both directions with a source (source-wave). However, the gait generation mechanism is still unclear because of the complex interaction between neural control and dynamic body systems through sensory information (embodied sensorimotor interaction). Our previous study showed that local sensory feedback has a function to generate the three interlimb coordination observed in multi-legged animals using a simple model. In this study, to further understand the functional role of sensory feedback, we investigate the effect of the sensory feedback on a three-dimensional multi-legged robot model developed. The simulation result with the ten-legged dynamic robot model shows that the sensory feedback also generates various wave gaits in the robot due to the embodied sensorimotor interaction. The generated gaits are not predetermined but emerge in a decentralized manner. Parts of generated gaits are similar to direct and retrograde wave gaits. In addition, sink wave gait, in which the swing movements sink in the center of the robot, is also observed.

[1]  K. Tsuchiya,et al.  Generation of Direct-, Retrograde-, and Source-Wave Gaits in Multi-Legged Locomotion in a Decentralized Manner via Embodied Sensorimotor Interaction , 2021, Frontiers in Neural Circuits.

[2]  Akio Ishiguro,et al.  Decoding the essential interplay between central and peripheral control in adaptive locomotion of amphibious centipedes , 2019, Scientific Reports.

[3]  Fumitoshi Matsuno,et al.  Simple analytical model reveals the functional role of embodied sensorimotor interaction in hexapod gaits , 2018, PloS one.

[4]  Akio Ishiguro,et al.  Decentralized control mechanism underlying interlimb coordination of millipedes , 2017, Bioinspiration & biomimetics.

[5]  Akio Ishiguro,et al.  Decentralized control scheme for myriapod robot inspired by adaptive and resilient centipede locomotion , 2017, PloS one.

[6]  A. Ishiguro,et al.  Common mechanics of mode switching in locomotion of limbless and legged animals , 2014, Journal of The Royal Society Interface.

[7]  Shinya Aoi,et al.  A stability-based mechanism for hysteresis in the walk–trot transition in quadruped locomotion , 2013, Journal of The Royal Society Interface.

[8]  D. Owaki,et al.  Simple robot suggests physical interlimb communication is essential for quadruped walking , 2013, Journal of The Royal Society Interface.

[9]  F. Wörgötter,et al.  Self-organized adaptation of a simple neural circuit enables complex robot behaviour , 2010, ArXiv.

[10]  Auke Jan Ijspeert,et al.  Central pattern generators for locomotion control in animals and robots: A review , 2008, Neural Networks.

[11]  Ansgar Büschges,et al.  Organizing network action for locomotion: Insights from studying insect walking , 2008, Brain Research Reviews.

[12]  F. Delcomyn Insect walking and robotics. , 2003, Annual review of entomology.

[13]  J. Schmitz,et al.  Rhythmic patterns in the thoracic nerve cord of the stick insect induced by pilocarpine , 1995, The Journal of experimental biology.