Reconfigurable curved beams for selectable swimming gaits in an underwater robot

Mohammad Sharifzadeh, Yuhao Jiang, Daniel M. Aukes

Research output: Contribution to journalArticlepeer-review

Abstract

Rowing is a swimming motion employed by a number of animals via tuned passive biomechanics and active gait strategies. This gait generates positive net thrust (or moment) by having a higher drag profile in the power stroke compared with the recovery stroke, which is obtained via faster actuation speed or higher effective area. In this letter, we show that using the preferential buckling of curved beams in swimming robots can, via a passive reduction of effective area in recovery stroke, be used to generate positive net thrust and moment. Additionally, these curved beams can be actively tuned to alter their behavior on demand for use in swimming applications, and can be used in an underwater robot to switch between rowing and flapping gaits. A dynamic model has been developed to model the swimming behavior of a robot using buckling joints. A design optimization has been carried out, using the Covariance Matrix Adaption Evolution Strategy (CMA-ES), to find the design and gait parameters that maximize the robot's forward swimming speed. A series of experimental gait searches have subsequently been conducted on the resulting optimal design, again using CMA-ES with the goal of finding the optimal gait pattern across a number of swimming strategies such as paddling, flapping, and undulation. By actively altering the curved beam's buckling limits, an untethered robot has been developed that maneuvers in water across each of these swimming strategies. The findings suggest that tuning the preferential buckling limits of curved beams can be an effective and potentially advantageous approach for producing directional thrust and moments.

Original languageEnglish (US)
Article number9369911
Pages (from-to)3437-3444
Number of pages8
JournalIEEE Robotics and Automation Letters
Volume6
Issue number2
DOIs
StatePublished - Apr 2021
Externally publishedYes

Keywords

  • And learning for soft robots
  • Compliant joints and mechanisms
  • Control
  • Modeling
  • Soft robot applicationsc
  • Soft robot materials and design

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Biomedical Engineering
  • Human-Computer Interaction
  • Mechanical Engineering
  • Computer Vision and Pattern Recognition
  • Computer Science Applications
  • Control and Optimization
  • Artificial Intelligence

Fingerprint Dive into the research topics of 'Reconfigurable curved beams for selectable swimming gaits in an underwater robot'. Together they form a unique fingerprint.

Cite this