Multi-arm Robotic Swimmer Actuated by Antagonistic SMA Springs

Michael Sfakiotakis, Asimina Kazakidi, Theodoros Evdaimon, Avgousta Chatzidaki, Dimitris Tsakiris

Research output: Contribution to conferencePaperpeer-review

Abstract

The present paper considers the modeling, control, development and experimental validation of bio-inspired multiarm underwater robotic swimmers actuated by compliant actuating elements, in the context of the soft robotics paradigm. Each one of the swimmer's compliant arms is actuated at its base by a pair of antagonistic compliant shape memory alloy (SMA) springs. The base joint of each such arm displays hysteretic behavior and asymmetries, which are compensated via the modified Prandtl-Ishlinskii (MPI) model of the joint's response, in conjunction with angular position feedback from a potentiometer. This closed-loop control scheme achieves fast and efficient tracking of a sculling joint motion profile. Experimental results based on a pair of such submerged arms, integrated in a catamaran hull, indicate the feasibility of this actuation and control scheme, providing propulsive speeds up to approximately 0.5 arm lengths per second (~50 mm/sec) and propulsive forces up to 30 mN. The experimental studies presented, regarding the effect of the arm kinematic parameters on propulsive speed and force, are in qualitative agreement with previous results of our group for rigid-actuator multi-arm underwater swimmers
Original languageEnglish
Pages1540-1545
DOIs
Publication statusPublished - Sept 2015
EventIntelligent Robots and Systems - Hamburg, Germany
Duration: 29 Sept 201502 Oct 2015

Conference

ConferenceIntelligent Robots and Systems
Abbreviated titleIROS-2015
Country/TerritoryGermany
CityHamburg
Period29 Sept 201502 Oct 2015

Keywords

  • bio-inspired robotics
  • soft robotics
  • marine robotics
  • shape memory alloys
  • control
  • Prandtl-Ishlinskii model

Fingerprint

Dive into the research topics of 'Multi-arm Robotic Swimmer Actuated by Antagonistic SMA Springs'. Together they form a unique fingerprint.

Cite this