This article, which is an extension of the previous work of the authors (Tanveer and Ahmad (2022)) on yaw dynamics, investigates the modeling and control of heave degree-of-freedom of a compact custom designed ROV. Wherein, nonlinear data-driven modeling strategy is adopted to develop a high-fidelity heave dynamic model. The proposed modeling approach uses open-loop real-time experimental data to derive a high-fidelity NARX model of the vehicle. The resulting model accommodates the dynamics of the system in addition to the tether dynamics. The advantage of this approach is its ability to eliminate the need for intricate controller tuning. The identified model consistently demonstrated fitness scores ranging from 82% to 92% in both self-validation and cross-validation tests conducted on distinct datasets. This relative advantage is exemplified in real-time through the testing of a Genetic Algorithm Proportional-Integral (GAPI) controller. The performance of GAPI is subsequently compared with the relatively recent Marine Predators Algorithm (MPA) and the more conventional root-locus tuned PI controllers. The experimental results demonstrate that GAPI provides the most favorable response, achieving a 35%, 76%, and 44% improvement in rise time, percent overshoot and peak time, respectively. Furthermore, the controller effort required by GAPI running on an ATmega328 chipset is 22% less than its counterparts.

Swimming in fish arises from tightly integrated neural, muscular, skeletal, and hydrodynamic processes that are difficult to capture in compact, transferable models for robotics. An interpretable system identification (SySID) is presented that bidirectionally maps between electromyography (EMG) and kinematics in freely swimming koi and further tests its generalization to a robotic fish. Synchronized EMG and kinematic are collected across laminar, Kármán vortex, and reverse Kármán vortex flows spanning 0.146–0.274 m s−1. A linear autoregressive with exogenous input (ARX) model architecture is chosen to capture both feedforward (EMG to kinematics) and feedback (kinematics to EMG) pathways, enabling the extraction of key system parameters, such as natural frequency, damping ratio, and input–output delays. Cross-individual validation demonstrates robust performance and identifies the best-performing fish-trained model, which is then evaluated for cross-domain transfer by replacing EMG input with processed pulse width modulation actuation signals from a robotic fish. Despite differences in mechanics and actuation physics, predictions closely match measured trajectories (mean R2 = 0.86 ± 0.13), substantially outperforming a deep neural network (97.8% higher percentage fit index) trained on the same biological datasets. These findings show that compact, interpretable SySID models enable accurate bio-to-robot transfer without robot-specific retraining, grounding robotic motion models directly in biological function rather than imitation.