Swimming with robots: investigating fish locomotion, sensing, and schooling behavior with robotic swimmers.

Fish display remarkable locomotor and social abilities, from efficient swimming to coordinated schooling, that have inspired the design of various robotic fish. While robotics has largely benefited from biology, fish-like robots are increasingly used as scientific tools to investigate fundamental questions in biomechanics, sensorimotor control, and collective behavior. This paper reviews how robotic models have been employed to study the neuromechanical basis of swimming, the role of sensory feedback in locomotion, and the mechanisms underlying social interactions in schools. We list open questions in biology and show that robotic approaches provide unique advantages to address them: they enable repeatable experiments, systematic variation of body and control parameters, and direct measurement of otherwise inaccessible quantities such as internal forces or energy use. A literature analysis reveals, however, that only a minority of robot-fish studies contribute to biological understanding, with most focusing on engineering design. Among biology-oriented studies, closed-loop robotic systems-capable of real-time adaptation-remain underrepresented but are essential for probing sensorimotor and social feedback mechanisms. We conclude by outlining future directions combining robotics, simulations, and emerging experimental technologies to unravel the multi-scale feedback loops that shape fish locomotion and schooling.