Google Meet link: meet.google.com/gde-trxz-vdu
Rodrigo Amaducci,
Irene Elices,
Manuel Reyes-Sanchez,
Alicia Garrido-Peña,
Rafael Levi,
Francisco B Rodriguez,
Pablo VaronaA hybrid robot or hybrot is a technology that combines living cells and networks with robotics. This technology is largely undeveloped and has been mainly implemented with neuron cultures and multichannel electrode arrays [1,2]. Hybrots have a lot of potential to study neural networks properties involved in the control of locomotion, sensorimotor transformation and behavior. Central pattern generators (CPG) are neural circuits that produce robust rhythmic sequences involved in motor functions such as breathing or walking. Because of their role in generating and coordinating motor rhythms, bio- inspired CPGs have been widely employed in robotic paradigms [3] including the design of novel mechanisms for autonomous locomotion [4]. However, the intrinsic mechanisms that give rise to the coordination of living CPG dynamics have not been used yet for hybrid robot implementation. In this work, we present the first hybrot controlled by a living CPG from the crab Carcinus Maenas. The robot and the living neural circuit are connected following a closed-loop protocol that involves a dynamic-clamp setup to communicate both elements through Bluetooth signaling. We show that effective robotic locomotion is achieved when it is controlled and coordinated by the flexible rhythmic sequences produced by the circuit of living motoneurons. The robot is equipped with a light sensor that sends a sensory feedback to the CPG in the form of intracellular current injection. We report the analysis of the presence of dynamical invariants in the intervals that build up the sequential activations of the living circuit [5] and how they are transmitted to the robot resulting in a coordinated locomotion. In turn, the robotic sensory feedback is translated into a variation of the living network activity while keeping the motor sequence, which results in a coherent response to the change in the environmental light.
Acknowledgements We acknowledge support from AEI/FEDER PGC2018-095895-B-I00 and TIN2017-84452-R.
References 1\. Potter SM. Hybrots: hybrid systems of cultured neurons+robots, for studying dynamic computation and learning. Proc 2002 Simul Adapt Behav 7 Work Mot Control Humans Robot Interplay Real Brains Artif Devices. Edinburgh, Scotland; 2002.
2\. Li Y, Sun R, Wang Y, Li H, Zheng X. A Novel Robot System Integrating Biological and Mechanical Intelligence Based on Dissociated Neural Network- Controlled Closed-Loop Environment. PLoS One. Public Library of Science; 2016;11:e0165600. https://doi.org/10.1371/journal.pone.0165600
3\. Ijspeert AJ. Central pattern generators for locomotion control in animals and robots: a review. Neural Netw. 2008;21:642–53. http://dx.doi.org/10.1016/j.neunet.2008.03.014
4\. Herrero-Carrón F, Rodríguez FB, Varona P. Bio-inspired design strategies for central pattern generator control in modular robotics. Bioinspiration and Biomimetics. 2011;6:16006. http://dx.doi.org/10.1088/1748-3182/6/1/016006
5\. Elices I, Levi R, Arroyo D, Rodriguez FB, Varona P. Robust dynamical invariants in sequential neural activity. Sci Rep. 2019;9:9048. https://doi.org/10.1038/s41598-019-44953-2