Design and testing of a selectively compliant underactuated hand

Abstract

Motivated by the requirements of mobile manipulation, a compliant underactuated hand, capable of locking individual joints, has been developed. Locking is accomplished with electrostatic brakes in the joints and significantly increases the maximum pullout forces for power grasps. In addition, by locking and unlocking joints, the hand can adopt configurations and grasp sequences that would otherwise require a fully actuated solution. Other features of the hand include an integrated sensing suite that uses a common transduction technology on flexible printed circuits for tactile and proprioceptive sensing. The hand is analyzed using a three-dimensional rigid body analysis package with efficient simulation of compliant mechanisms and contacts with friction. This package allows one to evaluate design tradeoffs among link lengths, required tendon tensions, spring stiffnesses and braking requirements to grasp and hold a wide range of objects. Results of grasping and pullout tests confirm the utility of the simulations.

Stiffness Control With Shape Memory Polymer in Underactuated Robotic Origamis

“…The geometrical parameters in (3), (4), and the following equations are depicted in Fig. 2 and their values are presented in Table I.…”

“…The need for lighter and easier to control robots has lead to an alternate approach of using under-actuated mechanisms for activating many DoFs using a single source of actuation [2], [3]. Moreover, the inherent tolerance of the under-actuated hands to impacts and their ability to conform to their environment through distribution of the input actuation between the joints [4] makes them soft and inherently safe for human interaction. Under-actuated robotic grippers can preform different grasping motions depending on the object shape and the contact points [5], [6].…”

“…tasks [4], [9], [10]. Switching between discrete modes of operation rather than fine control over the transmission ratio and requiring an additional source of actuation are two of the main limitations in the proposed methods.…”

Tactile Sensors for Friction Estimation and Incipient Slip Detection—Toward Dexterous Robotic Manipulation: A Review

“…Features relating to signal power and mean square coherence across all sensing element pairs were extracted from the BioTac [57], a sensor with an array of capacitive sensors and accelerometers [58,59], and a PVDF-based sensor with a curved, textured rubber, outer skin. Similar slip-state classification was performed by [60] on features extracted from three unidentified sensors using a Long-Short-Term Memory NN.…”

A Brief Survey on the Role of Dimensionality Reduction in Manipulation Learning and Control

“…Based on the studies of synergies, low-cost hands, compliant and underactuated, are becoming increasingly popular in the robotics community [80], [81], [82], [67]. On one side, those hands can simplify grasping and potentially in-hand manipulation.…”