BlueGuppy: tunable kinematics enables maneuverability in a minimalist fish-like robot.

Aquatic ecosystems vital to biodiversity and climate change-such as coral reefs, kelp forests, and mangrove forests-are often cluttered with natural obstacles. To navigate these complex habitats, fish have evolved relatively small body sizes and outstanding maneuverability. In contrast, most unmanned underwater vehicles currently deployed for ocean monitoring are bulky and slow, limiting their ability to access these environments. Developing small and agile underwater robots that mimic native fish species provides a unique opportunity for automated sampling of dynamic aquatic ecosystems. In this paper, we present BlueGuppy, a miniature, low-cost, and untethered fish-like robot (9.5×2.4×3.0cm, 33.1 g) capable of maneuvering with a single actuator. It achieves swimming speeds of up to 2.8 body lengths per second and can execute tight turns with small circles 1.4 body lengths in radius. BlueGuppy can generate a net thrust even in the presence of an incoming flow, but the flow field around BlueGuppy only mirrors that of biological organisms when it is free-swimming, underscoring the importance of untethered robots for biomimetic research. We explored the maneuverability of BlueGuppy by tuning its kinematics. By varying its flapping frequencies and temporal bias, BlueGuppy can access a wide range of speeds and turning curvatures. The combination of speed, maneuverability, and simplicity establishes BlueGuppy as a unique platform in the literature with tremendous potential for both uncovering the biomechanics of schooling fish and advancing the state-of-the-art in autonomous ocean sampling.