Investigating the computational principles of motor control

This project, which lies at the intersection of the physical sciences, neuroscience and engineering, aims to investigate the computational principles of motor control, combining theoretical methods from control engineering and a data-driven approach.

I will look more specifically look at the control of 3D rotations, a constituent of both arm reaching and orienting head movements, and will investigate the computational requirements for the execution of directed movements in 3D. Breaking down the 3D rotational control problem into its building blocks will hopefully provide a good basis for studying the algorithmic basis of motor control and forming intuitions for the mechanistic role of various brain circuits and their interactions.

I will initially consider the different ways in which rotations can be represented (e.g using quaternions or Euler angles) and - using a control-theoretic approach - study optimal control algorithms formulated under various known representations of rotations.

Furthermore, the existence of a close collaboration between our group in the Engineering department and a group in the MRC Laboratory for Molecular Biology will enable me to access an experimental setup which uses optogenetic techniques to study the neural circuits responsible for head movements in mice, and thus offer the opportunity to obtain data from experiments specifically targeted at the understanding of the control of 3D rotations. This will hopefully yield further insights into the computational principles of this motor behavior, and possibly help adjudicate between the various internal representations of rotations.

Overall, this project will involve theoretical tools taken from control theory and physics, and apply them to the problem of motor control, a key question in neuroscience, also closely connected to developments in other engineering fields such as robotics.