Stable Grip Control on Soft Objects With Time-Varying Stiffness

Humans can hold a live animal like a hamster without overly squeezing despite the fact that its soft body undergoes impedance and size variations due to breathing and wiggling. Although the exact nature of such biological motor controllers is not known, existing literature suggests that they maintai...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on robotics 2016-06, Vol.32 (3), p.626-637
Main Authors: Nanayakkara, Thrishantha, Jiang, Allen, del Rocio Armas Fernandez, Maria, Hongbin Liu, Althoefer, Kaspar, Bimbo, Joao
Format: Article
Language:English
Subjects:
Computer simulation
Controllers
Convergence
Dynamics
Failure
Force
Grippers
Impedance
Mathematical models
Probability
Reduction of probability of failure
Robotics
Robots
Simulation
Stability
Stability criteria
stable grip control
Uncertainty
variable stiffness objects
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Humans can hold a live animal like a hamster without overly squeezing despite the fact that its soft body undergoes impedance and size variations due to breathing and wiggling. Although the exact nature of such biological motor controllers is not known, existing literature suggests that they maintain metastable interactions with dynamic objects based on prediction rather than reaction. Most robotic gripper controllers find such tasks very challenging mainly due to hard constraints imposed on the stability of closed-loop control and inadequate rates of convergence of adaptive controller parameters. This paper presents experimental and numerical simulation results of a control law based on a relaxed stability criterion of reducing the probability of failure to maintain a stable grip on a soft object that undergoes temporal variations in its internal impedance. The proposed controller uses only three parameters to interpret the probability of failure estimated using a history of grip forces to adjust the grip on the dynamic object. Here, we demonstrate that the proposed controller can maintain smooth and stable grip tightening and relaxing when the object undergoes random impedance variations, compared with a reactive controller that involves a similar number of controller parameters.
ISSN:1552-3098
1941-0468
DOI:10.1109/TRO.2016.2549545