Inference of 6-DOF Robot Grasps using Point Cloud Data

Hsien-I Lin,Minh Nguyen Cong 제어로봇시스템학회 2019 제어로봇시스템학회 국제학술대회 논문집 Vol.2019 No.10

Recent decade, grasping pose perception for robot grasping field takes an important role in developing intelligent robotic applications. In this paper, we present a novel pipeline to infer grasp configuration for 3-finger gripper from point cloud data of a working environment. Most of the previous works are on 2D or 2.5D inputs to perform the 2D grasps. However, the 2D grasp configuration is not really compatible with the physical robot structure, the grasp successful rate on a robot system is only around 73% to 80%. To improve the robustness and computation, our goal is to propose an approach that can generate reliable grasping configurations in the cluster space. To accomplish this, we adopt a point cloud deep network to learn the shapes of objects and recognize their surfaces from the partial view. Afterwards, the grasp candidates are inferred based on the object models and reference grasping configuration sets. In our experiment, the proposed point cloud network had a better successful grasping rate compared to previous approaches in the context of varied and imbalanced data. And the grasping inference was highly viable in the complicated working environment.

A 1-Dof bidirectional graspable finger mechanism for robotic gripper

Hongliang Hua,Zhenqiang Liao,YongJiang Chen 대한기계학회 2020 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.34 No.11

The grasping flexibility and control simplicity are essential requirements for robotic gripper. Take above requirements into consideration, we present a 1-Dof four-bar finger mechanism which could produce bidirectional symmetric grasping motion and exhibit a bidirectional grasping ability like soft gripper. The finger mechanism could not only to grasp the object’s outer surface from outside to inside, but also able to extend itself from inside to outside and “grasp” the inner surface of the object. The effects of structural parameters of the finger mechanism on kinematic behavior have been investigated by kinematic modeling and simulations, and some design suggestions are given. A gripper prototype composed of two finger mechanisms is developed, and the effectiveness of the finger mechanism in bidirectional grasping is validated by grasping experiments. Results reveal that due to the symmetrical movement of the two finger mechanisms, they could apply a pair of symmetrical grasping force to the object and keep the stability of the grasping.

Detecting Graspable Rectangles of Objects in Robotic Grasping

Lu Chen,Panfeng Huang,Yuanhao Li,Zhongjie Meng 제어·로봇·시스템학회 2020 International Journal of Control, Automation, and Vol.18 No.5

Most convolutional neural network based grasp detection methods evaluate the predicted grasp by computing its overlap with the selected ground truth grasp. But for typical grasp datasets, not all graspable examples are labelled as ground truths. Hence, directly back propagating the generated loss during training could not fully reveal the graspable ability of the predicted grasp. In this paper, we integrate the grasp mapping mechanism with the convolutional neural network, and propose a multi-scale, multi-grasp detection model. First, we connect each labeled grasp and refine them by discarding inconsistent and redundant connections to form the grasp path. Then, the predicted grasp is mapped to the grasp path and the error between them is used for back-propagation as well as grasp evaluation. Last, they are combined into the multi-grasp detection framework to detect grasps with efficiency. Experimental results both on Cornell Grasping Dataset and real-world robotic grasping system verify the effectiveness of our proposed method. In addition, its detection accuracy keeps relatively stable even in the circumstance ofhigh Jaccard threshold.

Grasp Behavior Analysis Using Muscle and Postural Hand Synergies for Smartphones

Sung Hee Ahn,Sanghyun Kwon,Youngjin Na,Myunghwan Yun 한국정밀공학회 2021 International Journal of Precision Engineering and Vol.22 No.4

Smartphones are currently among the most common handheld devices. Previous studies on such handheld touchscreen devices focused on thumb operations or reach zones by measuring individual muscle or joint angles. However, they were limited to thumb operations and did not consider grasping. In this study, we investigated the grasp types of touchscreen devices and other objects included in an existing grasp taxonomy. To this end, principal component analysis and latent profile analysis clustering were used for extracting and grouping muscle and postural synergies. Fourteen healthy subjects performed up to 15 hand grasps, including that with a smartphone. Electromyography (EMG) data were measured on six muscles in the forearm and the hand, and joint angles were measured for 22 joints in the hand. The first two muscle synergies from the EMG data and the first three postural synergies from the kinematic data were found to account for over 60% of the overall grasping. In terms of the synergies, the grasp for handheld touchscreen devices showed unique characteristics in terms of muscle and postural synergies compared to other objects. The obtained results may aid in understanding of grasping behaviors for handheld touchscreen devices in various applications.

The Effect of Exercise Using Grasping on Shoulder Muscle Activity and Muscle Thickness of Patients with Rotator Cuff Repair

( Eun-young Choi ),( Hyun Choi ) 대한물리치료학회 2016 대한물리치료학회지 Vol.28 No.1

Purpose: This study intended to mediate shoulder exercise using grasping and investigate examine how its effects on affects activities of the shoulder surrounding muscles of the shoulder and thickness of shoulder muscle to suggest effective exercise for patients with rotator cuff repair to return to daily life. Methods: This study has been done targeteding male patients more than 6 weeks after having En-masse Suture Bridge Technique as rotator cuff repair of the right shoulder, aged 40 to 55 in a medical institution located in Jeollanam-do. Experimental group I (n=15) was selected for a conducting common exercise therapy program and Experimental group II (n=15) was selected for an conducting exercise therapy program using grasping. %RVC was measured by surface electromyography and muscle thickness was measured by ultrasound before mediation to for analysisze of before and after results within group and between groups. Results: In comparison of change of %RVC, Tthere were meaningful differences in only posterior deltoid and infraspinatus in comparing a change of %RVC within experimental group I (p<0.001),. and Tthere were meaningful differences in both supraspinatus and infraspinatus in muscle thickness within experimental group I (p<0.001). In comparison of change of %RVC, Tthere were meaningful differences in anterior deltoid, posterior deltoid and infraspinatus in comparing a change of %RVC within experimental group II (p<0.05)(p<0.01) (p<0.001), and there were meaningful differences in both supraspinatus and infraspinatus in muscle thickness within experimental group II (p<0.001). In comparison of change of %RVC between groups, Tthere was a meaningful difference in only posterior deltoid in comparing a change of %RVC between groups (p<0.001). Conclusion: We found that exercise therapy using grasping has a positive effect on shoulder stability muscles such as supraspinatus muscle and infraspinatus muscle and activity of deltoid muscle.

Benchmarking anthropomorphic hands through grasping simulations

Llop-Harillo Immaculada,Iserte José L,Pérez-González Antonio 한국CDE학회 2022 Journal of computational design and engineering Vol.9 No.2

In recent decades, the design of anthropomorphic hands has been developed greatly improving both cosmesis and functionality. Experimentation, simulation, and combined approaches have been used in the literature to assess the effect of design alternatives (DAs) on the final performance of artificial hands. However, establishing standard benchmarks for grasping and manipulation is a need recognized among the robotics community. Experimental approaches are costly, time consuming, and inconvenient in early design stages. Alternatively, computer simulation with the adaptation of metrics based on experimental benchmarks for anthropomorphic hands could be useful to evaluate and rank DAs. The aim of this study is to compare the anthropomorphism of the grasps performed with 28 DAs of the IMMA hand, developed by the authors, using either (i) the brute-force approach and grasp quality metrics proposed in previous works or (ii) a new simulation benchmark approach. The new methodology involves the generation of efficient grasp hypotheses and the definition of a new metric to assess stability and human likeness for the most frequently used grasp types in activities of daily living, pulp pinch and cylindrical grip, adapting the experimental Anthropomorphic Hand Assessment Protocol to the simulation environment. This new simulation benchmark, in contrast to the other approach, resulted in anthropomorphic and more realistic grasps for the expected use of the objects. Despite the inherent limitations of a simulation analysis, the benchmark proposed provides interesting results for selecting optimal DAs in order to perform stable and anthropomorphic grasps.

정상인의 완관절 신전, 굴곡, 握力 시 표면근전도 측정 연구

김석희,이경진,최유민,김주용,육태한,이상룡,김종욱 대한침구의학회 2015 대한침구의학회지 Vol.32 No.3

Purpose : This study aimed to prove that surface electromyography(SEMG) can be used to identify the degree of symptoms of diseases in the upper extremities; it also aimed to confirm various potential therapeutic effects through an inquiry into the value measured by the SEMG on certain acupuncture points in the upper extremities. Methods : Fifty healthy people received instructions for the method of exercise: wrist flexion, extension and hand grasping. Disposable electrodes were attached to acupuncture points of Susamni(LI10), Naegwan(PC6), Oegwan(TE5) and below Sohae(HT3) two cun on both sides in flexion, extension and grasping to measure the SEMG values. The research results were analyzed using SPSS statistics Ver. 22.0(IBM, USA). Results : The average value was highest on Naegwan(PC6) in grasping, and the average SEMG value was higher in the order of grasping, extension and flexion. The asymmetry index of each point was Susamni(LI10) 16.26±13.59 %, Oegwan(TE5) 20.38±15.59 %, below Sohae(HT3) two cun 20.89±16.77 %, Naegwan(PC6) 22.49±14.91 % in wrist extension, Susamni(LI10) 25.99± 21.44 %, Oegwan(TE5) 21.15±15.94 %, below Sohae(HT3) two cun 19.62±15.46 %, Naegwan(PC6) 19.93±16.85 % in wrist flexion, Susamni(LI10) 16.60±12.21 %, Oegwan(TE5) 10.94±8.29 %, below Sohae(HT3) two cun 15.20±12.60 %, Naegwan(PC6) 11.68±7.77 % in grasping. Conclusions : In this study, to identify the degree of symptoms of diseases in the upper extremities and confirm therapeutic effects, it is necessary to analyze the calculated percentage and compare the SEMG measurement of special points with those of other points, and with the asymmetry index.

Fast suction-grasp-difficulty estimation for high throughput plastic-waste sorting

Sangwoo Um,김경수,김수현 대한기계학회 2023 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.37 No.2

The selection of the grasping location is the most important task for robots that handle randomly shaped objects. In previous studies, the grasp quality was accurately evaluated, but the speed was much too low for high-throughput applications, and the focus was mainly on industrial products. In this study, a large-scale dataset for randomly deformed plastics is constructed. We propose a contact-area estimation model and difficulty function for a quantitative analysis of surface conditions. Synthetic labels were calculated using the tuned difficulty function for donut-shaped contact areas. We trained the network containing a pretrained encoder and decoder with skip connections for grasp-difficulty map estimation. Graspdifficulty estimations for multiple objects required at most 30.9 ms with an average error rate of 1.65 %. The algorithm had a 94.4 % grasping success rate and its computational efficiency was compared with that in previous studies. The algorithm enables the rapid sorting of continuously conveyed objects with higher throughput.

Soft grasping mechanisms composed of shape memory polymer based self-bending units

Wang, Wei,Yu, Chak Yuk,Abrego Serrano, Pablo Antonio,Ahn, Sung-Hoon Elsevier 2019 Composites Part B, Engineering Vol.164 No.-

<P><B>Abstract</B></P> <P>Recent advances in four-dimensional (4D) printing technology have developed various approaches for creating shape-morphing structures and devices. However, it is impractical as a universal method to fabricate directly available shape-morphing structures in any desired shapes. These 4D printed shape-morphing structures are either limited in structures with flat architectures or needed for the subsequent tedious thermo-mechanical programming process. To overcome these limitations, this work, as an example, demonstrates a versatile fabrication method for constructing soft grasping mechanisms with different architectures via mechanically assembling the basic self-bending units made of polymer-paper bilayer composite. The bilayer is fabricated in a single printing process by three-dimensional printing polymer fibers in parallel and layer-by-layer on the paper substrate. All the basic units are cut out in strip shape from the same piece of bilayer along the direction of the printed fiber being able to generate bending deformation in the direction transverse to the printed fibers upon heating. The obtained units are then mechanically assembled using pre-fabricated hard components for constructing three grasping mechanisms with different architectures. Upon heating, each unit generating the bending deformation endows the assembled structure with a desired integral shape-transition for the function of grasping. This method to the mechanical assembly of self-bending units provides new routes to the design of polymer-based shape-morphing structures.</P>

The Effect of Exercise Using Grasping on Shoulder Muscle Activity and Muscle Thickness of Patients with Rotator Cuff Repair

Choi, Eun-Young,Choi, Hyun 대한물리치료학회 2016 대한물리치료학회지 Vol.28 No.1

Purpose: This study intended to mediate shoulder exercise using grasping and investigate examine how its effects on affects activities of the shoulder surrounding muscles of the shoulder and thickness of shoulder muscle to suggest effective exercise for patients with rotator cuff repair to return to daily life. Methods: This study has been done targeteding male patients more than 6 weeks after having En-masse Suture Bridge Technique as rotator cuff repair of the right shoulder, aged 40 to 55 in a medical institution located in Jeollanam-do. Experimental group I (n=15) was selected for a conducting common exercise therapy program and Experimental group II (n=15) was selected for an conducting exercise therapy program using grasping. %RVC was measured by surface electromyography and muscle thickness was measured by ultrasound before mediation to for analysisze of before and after results within group and between groups. Results: In comparison of change of %RVC, Tthere were meaningful differences in only posterior deltoid and infraspinatus in comparing a change of %RVC within experimental group I (p<0.001),. and Tthere were meaningful differences in both supraspinatus and infraspinatus in muscle thickness within experimental group I (p<0.001). In comparison of change of %RVC, Tthere were meaningful differences in anterior deltoid, posterior deltoid and infraspinatus in comparing a change of %RVC within experimental group II (p<0.05)(p<0.01) (p<0.001), and there were meaningful differences in both supraspinatus and infraspinatus in muscle thickness within experimental group II (p<0.001). In comparison of change of %RVC between groups, Tthere was a meaningful difference in only posterior deltoid in comparing a change of %RVC between groups (p<0.001). Conclusion: We found that exercise therapy using grasping has a positive effect on shoulder stability muscles such as supraspinatus muscle and infraspinatus muscle and activity of deltoid muscle.