The overall scientific objective of the PlasticBrain project is to develop a deeper understanding of human motor
control capabilities and to apply them to the design and test of bio-inspired control algorithms for
anthropomorphic, as well as non-anthropomorphic, devices. The so-called synthetic approach will be employed
to investigate the key properties of the human motor system, namely redundancy and adaptability. The results of
such a primary investigation will then be applied to the design and test of new bio-inspired control algorithms
aiming to improve robustness, reconfigurability and adaptability in robotics, which can find natural application
in rehabilitation, prosthetics, brain-machine interface and intelligent devices. It is known that sensorimotor
coordination in humans and animals is performed via a multiplicity of control systems that act opportunistically
and synergistically depending on the context, on body configuration, on perceptual possibilities, etc. On the other
hand, robot controllers, and control systems in general, suffer from a significant lack of flexibility and
adaptability to large variations in configuration. To address these issues the project will conduct a solid
investigation founded on empirical research on the structure of neural motor control using state-of-the-art brain
science methods. This study will focus on reaching movements generalized to include the use of any body part as
potential effector in a global setting where whole body movements might be required. By looking at a better
understanding of biological motor control, we will simultaneously explore the development of a full-fledged
robotic control system that imitates biological reality. We will develop new control methods, based on Port
Hamiltonian formalism and integrate them within an artificial neuro-muscular architecture capable of handling
hyper-redundant structure in an adaptive and reconfigurable way. To validate the algorithms, we will use a
humanoid robotic platform, the iCub, a 53 degree of freedom anthropomorphic robot, and a nonanthropomorphic
platform shaped as an octopus tentacle, to ensure the validity of the bio-inspired control
algorithms on a general robotic platform.