Undulatory swimming is among nature's most efficient and widespread aquatic propulsion strategies. Eels demonstrate remarkably high swimming efficiency in their migratory journeys as they travel for thousands of kilometres without eating. Snakes display swimming adaptability regardless of their natural habitat. These swimmers generate propulsion by propagating a travelling wave along their bodies. The notable swimming performance depends on both the body travelling-wave kinematics and the surrounding hydrodynamic environment. The role of individual kinematic parameters and environmental effects remains difficult to systematically isolate in live animals. This thesis addresses this challenge using bio-inspired robotic swimmers as controlled experimental models for studying aquatic locomotion. Two geometrically similar versions of an anguilliform robotic swimmer 1-guilla were developed for free-swimming experiments in a pool and tethered experiments in a water channel.

First, we characterise the swimming performance of our robot as we vary its kinematic parameters, including joint amplitude, body wavelength, and frequency.
We identify a trade-off between speed and efficiency in free swimming. Speed, in terms of stride length, increases for increasing maximum tail angle, described by the proposed specific tail amplitude and reaches a maximum value around the specific tail amplitude of unity. Efficiency, in terms of the cost of transport, is affected by the whole-body motion. Cost of transport decreases for increasing travelling wave-like. The amplitude profile evolution along the length of the robot affects the performance. Increasing amplitude profiles reduce the energetic cost of swimming, at the expense of decreased maximum achievable swimming speed compared to constant amplitude profiles. Experimental constraints like tethering and depth of submergence affect the performance. Submerged swimming improves both speed and efficiency compared to surface swimming. Tethering reduces energetic cost by suppressing uncontrolled body roll.

We then investigate adaptive locomotion in vortical environments. The robotic swimmer is placed in the wake of an oscillating wing, generating a controlled vortex street. The controller of the robot is equipped with local curvature-based stretch feedback. The feedback mechanism promotes both frequency and phase synchronisation between the swimmer and the incoming vortices, and the energy consumption of the swimmer is reduced. Comparisons with zebrafish experiments and spiking neural network simulations further support the role of proprioceptive feedback as a biologically plausible mechanism for flow-mediated locomotor coordination.

Finally, we explore the consequences of ecological lifestyle adaptation on the swimming performance of snakes. A large dataset comprising more than \num{3500} swimming trials across diverse snake lifestyles is analysed using modal decomposition techniques to extract representative swimming patterns. These kinematics are replicated on the robotic swimmer and compared with optimised kinematics from a multi-objective experimental optimisation. Both kinematic and performance differences exist across lifestyles, with aquatic-proximity snakes resulting in higher swimming efficiency. Our results suggest that ecological adaptation has measurable signatures on locomotor performance and demonstrate how robotic experiments can bridge evolutionary biology and hydrodynamics.