Combining data and first principles in the synergy-based control of soft robotic

The long term vision of our research is that of fully amorphic robotic systems that are made from hundreds or thousands of soft, fluid-driven actuators and that can shape-shift in their entirety to perform a certain task.  In this project, we focus on the challenge of modelling and robustly controlling such hyper-actuated robotic systems by employing a small number of independent synergies that physically encode distributed behaviors within the robot. This approach inherently reduces the dimensionality of the modelling and control problem, which we seek to solve through a combination of first-principle and data-driven approaches.