Unmanned bio-robotic systems swimming in high-energy environments
The objective of this work is two fold:
- To understand and extract the biological principles of sensing, control and biomechanics that enable pinnipeds to swim and maneuver effectively in high-energy surf zone conditions.
- To adapt and demonstrate these principles on a freely swimming robotic platform that is able to take advantage of, and react to, multidirectional flows in surf zones.
Bio-robotic sea lion (Nicholas Marcouiller)
A bio-robotic sea lion platform (SEAMOUR) has been developed to investigate the biomechanics, sensing, and control strategies that enable aquatic animals to swim and maneuver well in high-energy aquatic environments. Inspired by the distributed control surfaces of California sea lions, the robot features an articulated body with actively controlled segments designed to generate turning moments and maintain stability during swimming.
The platform serves as a physical testbed for validating physics-based numerical models and evaluating closed-loop control strategies. Through controlled pool experiments, the bio-robotic sea lion enables systematic study of coordinated body actuation, sensor feedback, and maneuverability, with the long-term goal of achieving robust operation in surf-zone–like conditions.
Numerical model of the bio-robotic sea lion (Shraman Kadapa)
A high-fidelity numerical model of the bio-robotic sea lion has been developed to dynamically analyze and simulate six-degree-of-freedom body motions in a fluid environment. The equations of motion for the full articulated system are derived in closed form using an Euler–Poincaré formulation and solved symbolically.
The model captures both internal dynamics, including body inertia and joint torques, and external hydrodynamic effects, such as drag and added mass, for all body segments. This physics-based formulation provides a structured framework for studying system behavior and is particularly well suited for applications in control design, stability analysis, and reinforcement learning.