A soft actuator with tunable mechanical configurations for object grasping based on sensory feedback

  • Xi Fang
  • , Zemin Liu
  • , Yufei Hao
  • , Hui Yang
  • , Jiaqi Liu
  • , Zhexin Xie
  • , Li Wen*
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper, we propose a soft actuator embedded with conductive liquid alloy (eGaIn) and Shape Memory Polymer (SMP). The conductive liquid alloy functions as bending sensor as well as pressure sensor. The three segments of the SMP layer can be heated optionally by applying 400mA current to different sections of the conductive silver fiber to achieve tunable stiffness, and enable the soft actuator with variable degree-of-freedom (DoF) configurations. Through real-time feedback of the bending sensor that has been calibrated, we realized proprioception of all mechanical configurations of the soft actuator under free load. We also achieved geometrical feature recognition of different objects while all segments of the actuator were softened. Finally, experimental verification of the prototype was conducted on a variety of objects in condition that the geometrical features were unknown. The results show improved enveloping and reliable grasping performance.

Original languageEnglish
Title of host publicationRoboSoft 2019 - 2019 IEEE International Conference on Soft Robotics
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages25-30
Number of pages6
ISBN (Electronic)9781538692608
DOIs
StatePublished - 24 May 2019
Event2019 IEEE International Conference on Soft Robotics, RoboSoft 2019 - Seoul, Korea, Republic of
Duration: 14 Apr 201918 Apr 2019

Publication series

NameRoboSoft 2019 - 2019 IEEE International Conference on Soft Robotics

Conference

Conference2019 IEEE International Conference on Soft Robotics, RoboSoft 2019
Country/TerritoryKorea, Republic of
CitySeoul
Period14/04/1918/04/19

Fingerprint

Dive into the research topics of 'A soft actuator with tunable mechanical configurations for object grasping based on sensory feedback'. Together they form a unique fingerprint.

Cite this