An artificial cerebellum for robots

An international team of scientists, including those who are researchers at the University of Granada, have designed an artificial cerebellum for robots capable of mimicking the way to manipulate objects of human beings . The research was conducted over 4 years, framed in two European projects (SENSOPAC and REALNET).

The new model of artificial cerebellum allows robots to "learn" the intrinsic characteristics of an object that is presented (mass, inertia , resistance to the movement) and associate with another set of random features (color or shape) that help to distinguish it from other objects and to carry out a more precise handling. That is, artificially mimics how the nervous system leads to precisely the realization of movement and obtains knowledge of the objects it manages.

In the medium term, this study will create a new generation of robots that can interact with humans much more secure than at present. Furthermore, this type of study may help develop new treatments for diseases related to the cerebellum (as in the case of ataxia), as well as new methods of rehabilitation and prosthetics "smart."