Water Surface Swimming Dynamics in A Lightweight Undulatory Robot

Locomotion at the water surface occurs across scales, yet little is known about how elongate invertebrates like centipedes navigate such environments. Our previous research revealed that the centipede L. forficatus (L=2.3±0.3 cm) propels itself on the water surface via self-deformations consisting of waves of body bending which propagate in the direction of motion. A resistive force theory (RFT) approach, in which thrust and drag forces integrated over the body and maintained in force balance, predicted optimal swimming waves. To more systematically and comprehensively test surface RFT predictions, we developed a lightweight (46 g) undulatory surface swimmer robophysical model, actuated only by a single motor with a passively elastic buoyant tail (16-40 cm). The motor and a 3D printed linkage system drove the tail by generating linear oscillation at the base perpendicular to the direction of locomotion. The assembly was housed in a buoyant circularly shaped head. We varied the tail length and temporal frequency and tested locomotor performance for the emergent body kinematics. The RFT model using new surface element drag measurements captured swimming speed (5.5±1.1 cm/s) at larger BL and higher temporal frequency) indicating that RFT can be used at fluid surfaces in scenarios where robots and organisms do not coast significantly during cycles of self-deformation.