Interfacing neural and artificial systems: from neuroengineering to neurorobotics

Abstract

Neurosciences and robotics are coming close to each other in various ways: neuroscientists are developing increasingly sophisticated models of sensorimotor control and perception in animals and humans; roboticists are developing machines that incorporate quite advanced theoretical models and technologies. The two disciplines are complementary: robotics can take advantage of the new models developed by neuroscientists to develop new more performant biomimetic machines, and neuroscientists can use these robots to test their models of brain functions. Furthermore, neural systems and robotic devices can be interfaced, as in new generations of prosthetic limbs currently being investigated in various laboratories. The paper analyzes the motivations that are leading to the convergence between neurosciences and robotics, presents some examples of practical implementation, and discusses opportunities and problems posed by this new area of research closely related to neural engineering.