Hybrid sensorimotor performance
Contrôle sensorimoteur hybride
Our team uses hybrid systems, which mix biological control with artificial devices, in order to (i) increase our understanding of sensorimotor control and (ii) exploit this knowledge to restore and optimize movement. Instead of being pre-programmed in the brain, movement coordination largely depends upon multiple feedback loops that operate at different levels of the sensorimotor control system. For instance, muscle mechanics provides an instantaneous functional response to small perturbations, while segmental and transcortical reflexes are able to absorb increasingly larger perturbations by precisely coordinating muscle responses for the complex musculoskeletal design of our limbs. These loops are typically violated in the case of artificial devices, such as with prosthetic limbs whose biomechanics differs from that of original limbs, and which lack sensory feedback. We use a range of closed loop hybrid systems to investigate how these lower feedback loops interact with limb’s biomechanics, how they contribute to the production of normal, coordinated movements, and how to improve the design of hybrid control strategies to restore movements.
- Perceptually−guided deep neural networks for ego−action prediction: Object grasping.. Gonzalez−Diaz I, Benois−Pineau J, Domenger J−P, Cattaert D, de Rugy A. Pattern Recognition. 2019-04. 88 : 223-235.
- Reachy, a 3D-Printed Human-Like Robotic Arm as a Testbed for Human-Robot Control Strategies.. Sébastien Mick, Mattieu Lapeyre, Pierre Rouanet, Christophe Halgand, Jenny Benois-Pineau, Florent Paclet, Daniel Cattaert, Pierre-Yves Oudeyer, Aymar de Rugy. Front. Neurorobot.. 2019-08-14. 13
- Model and experiments to optimize co-adaptation in a simplified myoelectric control system.. M Couraud, D Cattaert, F Paclet, P Y Oudeyer, A de Rugy. J. Neural Eng.. 2018-01-30. 15 (2) : 026006.
- Distinct coordinate systems for adaptations of movement direction and extent.. Eugene Poh, Timothy J. Carroll, Aymar de Rugy. Journal of Neurophysiology. 2017-11-01. 118 (5) : 2670-2686.
- Action history influences subsequent movement via two distinct processes.. Welber Marinovic, Eugene Poh, Aymar de Rugy, Timothy J Carroll. eLife. 2017-10-23. 6