Algorithmic Fingertip Repositioning for Enhanced In-Hand Manipulation of the Objects

This paper focuses on a method to relocate the robotic fingertips on the surface of the object when the fingertips instantaneously hold the object under precision grasp. Precision grasp involves holding the object using fingertips. Finger gaiting involves repositioning the fingertips on the surface of the object and then manipulation of the object. During repositioning, one contact point leaves the object surface and recontacts at the other point. A metric is defined on the set of feasible grasp configurations to limit deviation from force closure during repositioning of the fingertips. Then, a manipulability based metric is described to search for the optimal goal grasp states on the object's surface. The manipulability based metric is used to search the grasp state to relocate the contacts, such that the range of object motion is increased.

Sensing form - finger gaiting as key to tactile object exploration - a data glove analysis of a prototypical daily task

Background Motor hand skill and associated dexterity is important for meeting the challenges of daily activity and an important resource post-stroke. In this context, the present study investigated the finger movements of right-handed subjects during tactile manipulation of a cuboid, a prototypical task underlying tactile exploration. During one motor act, the thumb and fingers of one hand surround the cuboid in a continuous and regular manner. While the object is moved by the guiding thumb, the opposed supporting fingers are replaced once they reach their joint limits by free fingers, a mechanism termed finger gaiting. Methods For both hands of 22 subjects, we acquired the time series of consecutive manipulations of a cuboid at a frequency of 1 Hz, using a digital data glove consisting of 29 sensors. Using principle component analysis, we decomposed the short action into motor patterns related to successive manipulations of the cuboid. The components purport to represent changing grasp configurations involving the stabilizing fingers and guiding thumb. The temporal features of the components permits testing whether the distinct configurations occur at the frequency of 1 Hz, i.e. within the time window of 1 s, and, thus, taxonomic classification of the manipulation as finger gaiting. Results The fraction of variance described by the principal components indicated that three components described the salient features of the single motor acts for each hand. Striking in the finger patterns was the prominent and varying roles of the MCP and PIP joints of the fingers, and the CMC joint of the thumb. An important aspect of the three components was their representation of distinct finger configurations within the same motor act. Principal component and graph theory analysis confirmed modular, functionally synchronous action of the involved joints. The computation of finger trajectories in one subject illustrated the workspace of the task, which differed for the right and left hands. Conclusion In this task one complex motor act of 1 s duration could be described by three elementary and hierarchically ordered grasp configurations occurring at the prescribed frequency of 1 Hz. Therefore, these configurations represent finger gaiting, described until now only in artificial systems, as the principal mechanism underlying this prototypical task. Trial registration clinicaltrials.gov, NCT02865642, registered 12 August 2016.

Sensing form - finger gaiting as key to tactile object exploration: A data glove analysis of a prototypical daily task

Background Motor hand skill with associated dexterity is an important facility in meeting the challenges of daily activity and an important resource post-stroke. In this context, the present study investigated the finger movements of right-handed subjects during tactile manipulation of a cuboid, a prototypical component of tactile exploration. Methods For both hands of 22 subjects, we acquired the time series of consecutive multifinger cuboid manipulations using a digital data glove consisting of 29 sensors. Of these, 16 recorded the bending of metacarpo-phalangeal (MCP) and proximal interphalangeal (PIP) joints of the fingers, MCP and interphalangeal IP joints of the thumb, palm arch and carpo-metacarpal (CMC) joint of the thumb, and abduction between fingers. Results Using principle component analysis we decomposed the short action into motor patterns related to successive manipulations of the cuboid. The fraction of variance described by the principal components indicated that three components described the salient features of the single motor acts for each hand. Striking in the finger patterns was the prominent and varying roles of the MCP and PIP joints of the fingers, and the CMC joint of the thumb. An important aspect of the three components was their representation of distinct finger configurations within the same motor act. Principal component and graph theory analysis confirmed modular, functionally synchronous action of the involved joints. The computation of finger trajectories in one subject illustrated the workspace of the task, which differed for the right and left hands. Conclusion The study substantiates finger gaiting, described until now only in artificial systems, as the principal mechanism underlying this prototypical task, which is ubiquitous in daily object shape recognition.

Learning dexterous in-hand manipulation

We use reinforcement learning (RL) to learn dexterous in-hand manipulation policies that can perform vision-based object reorientation on a physical Shadow Dexterous Hand. The training is performed in a simulated environment in which we randomize many of the physical properties of the system such as friction coefficients and an object’s appearance. Our policies transfer to the physical robot despite being trained entirely in simulation. Our method does not rely on any human demonstrations, but many behaviors found in human manipulation emerge naturally, including finger gaiting, multi-finger coordination, and the controlled use of gravity. Our results were obtained using the same distributed RL system that was used to train OpenAI Five. We also include a video of our results: https://youtu.be/jwSbzNHGflM .

Design, Analysis, and Testing of a Metamorphic Gripper

This paper introduces a novel metamorphic gripper proposed as a step towards the solution of the bin picking problem. This gripper makes uses of multiple poses while maintaining contact with the part allowing for in-hand manipulation without the use of finger gaiting. The gripper is analyzed using classical techniques of degrees of freedom and adjacency matrices. This analysis allows for a basic understanding of the motion of the gripper.

Novel Design of a Robotic Gripper Allowing for In-Hand Manipulation

The design of robotic end effectors can be loosely classified into two different types: complex anthropomorphic hands which allow for manipulation and simple open/close grippers which do not. This article investigates a design of a simple, industrial feasible end effector that allows for in-hand manipulation of parts. This end effector can be utilized in conjunction with a vision system to eliminate parts feeders and be able to pick parts straight from bins. The design utilizes passive joints that, when in a particular configuration, align (a self-motion singularity) to allow in-hand manipulation without regrasping or finger gaiting. A prototype end effector was fabricated and tested to prove the concept.