Robophysics of a Tensegrity Robotic Fish: Effects of Stiffness, Control, and Elastic Recoil on Swimming Efficiency

We present a tensegrity-based robotic fish that integrates compliant mechanics and bio-inspired control for robophysical studies of aquatic locomotion. The robot's body is made from serially connecting several identical modules. Each module has two rigid plates on two ends, which are connected by a tensegrity structure with both elastic tendons and rigid rods. Two motors at one end can bend the module through cables. Multiple modules are connected to form a compliant fish-like backbone, with electronics, drivers, and battery housed in a 3D-printed head and a flexible TPU tail providing thrust. The entire body will be enclosed in a waterproof skin. The design allows us to modify the stiffness of the body and fin spatially, enabling investigation of how their compliance affect swimming efficiency. Motion is generated by a two-neuron central pattern generator (CPG) that produces bending waves through differential motor actuation. We plan to explore alternative control methods. We also plan experiments to quantify swimming performance and energy efficiency as functions of body and fin stiffness as well as elastic energy usage. This robotic fish platform bridges soft robotics and physics, advancing understanding of compliant, energy-efficient underwater propulsion.