AquaClimber: a limbed swimming and climbing robot based on reduced order models

Area of Science:

• Robotics
• Biomimetics
• Fluid Dynamics

Background:

• Many legged robots mimic animal locomotion for running, jumping, and climbing.
• Extending limbed locomotion to swimming presents unique challenges.
• Few robots integrate climbing and swimming capabilities.

Purpose of the Study:

• To extend a reduced-order dynamic model for climbing locomotion to aquatic environments.
• To develop a multi-modal robot capable of both swimming and climbing.
• To investigate the effects of design parameters on swimming performance.

Main Methods:

• Developed a lateral plane reduced-order dynamic model for multi-modal locomotion.
• Designed and built the AquaClimber robot, inspired by human freestyle swimming.
• Validated simulation predictions against model results for swimming and climbing.

Main Results:

• The AquaClimber robot successfully demonstrated both swimming and vertical wall climbing.
• The reduced-order dynamic model accurately predicted swimming speed and behavior.
• Key parameters like hand design, arm compliance, and driving frequency were shown to influence swimming performance.

Conclusions:

• Successfully developed the first limbed robot with integrated aquatic and scansorial (climbing) multi-modal locomotion.
• The developed dynamic model provides a predictive tool for optimizing limbed swimming robots.
• This work opens new avenues for robots operating in complex, multi-domain environments.