salamander; central pattern generators; Robotics; spinal cord; Control of locomotion
Ramdya Pavan, Thandiackal Robin, Cherney Raphael, Asselborn Thibault, Benton Richard, Ijspeert Auke Jan, Floreano Dario (2017), Climbing favours the tripod gait over alternative faster insect gaits., in
Nature communications, 8, 14494-14494.
Karakasiliotis K., Thandiackal R., Melo K., Horvat T., Mahabadi N. K., Tsitkov S., Cabelguen J. M., Ijspeert A. J. (2016), From cineradiography to biorobots: an approach for designing robots to emulate and study animal locomotion, in
Journal of The Royal Society Interface, 13(119), 20151089-20151089.
Ryczko Dimitri, Thandiackal Robin, Ijspeert Auke J. (2016), Interfacing a salamander brain with a salamander-like robot: Control of speed and direction with calcium signals from brainstem reticulospinal neurons, in
2016 6th IEEE International Conference on Biomedical Robotics and Biomechatronics (BioRob), Singapore, Singapore.
The goal of this project is two-fold: (1) to use an interdisciplinary approach to decode the mechanisms of gait generation and gait transition in the salamander, and (2) to design a novel experimental setup that creates an interface between the salamander central nervous system and a salamander robot capable of swimming and walking. The focus is on the locomotor circuits in the brain stem and the spinal cord, in particular on decoding the interplay of descending control and spinal rhythm generation in locomotor activities. The novel experimental setup can be viewed as a neuroprosthetic device for a spinalized salamander, i.e. an animal with a complete lesion of the spinal cord.Using an interdisciplinary approach that combines neurophysiology, numerical simulations of coupled oscillators, and robotics, we will address the following questions:1. How do various descending pathways and spinal circuits interact to generate rich motor behavior?2. How closely can we replicate animals kinematic and electromyographic (EMG) data with a numerical model of the locomotor circuit and a robot?3. Can we record and/or stimulate descending pathways and use the recordings and/or stimulation patterns to steer a salamander robot equipped with a numerical model of the spinal circuits?These important questions, which are relevant for all tetrapods, will be addressed with a back and forth interaction between modeling, simulation, and experimentation. Research will be divided into two interconnected parts: (A) modeling, and (B) design and testing of the novel experimental setup. The models of the locomotor neural networks (Part A) will be based on systems of coupled nonlinear oscillators representing the central pattern generator circuits of the salamander spinal cord. These models are abstract enough to be tractable while being powerful explanatory and predictive tools, as illustrated in preliminary results. In order to investigate the feedback loops between the central nervous system, the body and the environment, these neural network models will be bidirectionally coupled with a novel salamander-like amphibious robot currently under construction. Part B, the design and testing of the novel experimental setup, will involve several steps: finalizing the new robot, interacting with neurophysiologists to prepare for (real-time) interfacing between neural recordings/stimulation and the robot, and performing a series of experiments where signals from the descending pathways are recorded (and possibly induced by electro-stimulation) and used to steer a salamander robot equipped with a model of the salamander spinal cord circuits.The expected impact of this project is two-fold: (1) a better understanding of the functioning of the spinal cord and of the descending pathways during locomotion in vertebrates, and (2) a novel experimental setup that creates an interface between the salamander central nervous system and a salamander robot. The novel setup will be greatly useful on one hand to address engineering problems related to neuroprostheses and interfacing a lower vertebrate nervous system with a robot, and on the other hand to test various hypotheses about sensorimotor loops in vertebrate animals. In the long term, the knowledge gathered in this project will hopefully contribute to designing therapies and/or neuroprosthetic devices for patients with spinal cord injuries (SCIs). In the short term, this study will significantly enhance our understanding of locomotor circuits in salamander and demonstrate how robotics can be used as a tool in neuroscience. Furthermore, since salamanders have capabilities of spinal regeneration and locomotor recovery after SCI that are quite unique among vertebrates, understanding the mechanisms of intact locomotion is essential to be able to properly characterize how locomotor function is recovered.