Spatial grasp synthesis for complex objects using model‐based simulation

Liu, Yanbin; Starr, Gregory P.; Wood, John; Lumia, Ronald

doi:10.1108/01439910510573255

Cited by 7 publications

(2 citation statements)

References 20 publications

(24 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While GraspIt! uses some of the grasp measures mentioned above to predict grasp success in the real world (also see [17] where objects with varied mass distributions were considered), some approaches use heuristic grasp measures such as hand grasp volume and hand symmetry [5] to quantify grasp quality on a physical robot. In addition, some approaches distinguish "contact robustness" from "grasp robustness" for grasp synthesis [24].…”

Section: Introductionmentioning

confidence: 99%

Human-guided grasp measures improve grasp robustness on physical robot

Balasubramanian

Brook

et al. 2010

2010 IEEE International Conference on Robotics and Automation

View full text Add to dashboard Cite

Humans are adept at grasping different objects robustly for different tasks. Robotic grasping has made significant progress, but still has not reached the level of robustness or versatility shown by human grasping. It would be useful to understand what parameters (called grasp measures) humans optimize as they grasp objects, how these grasp measures are varied for different tasks, and whether they can be applied to physical robots to improve their robustness and versatility. This paper demonstrates a new way to gather human-guided grasp measures from a human interacting haptically with a robotic arm and hand. The results revealed that a human-guided strategy provided grasps with higher robustness on a physical robot even under a vigorous shaking test (91%) when compared with a state-of-the-art automated grasp synthesis algorithm (77%). Furthermore, orthogonality of wrist orientation was identified as a key human-guided grasp measure, and using it along with an automated grasp synthesis algorithm improved the automated algorithm's results dramatically (77% to 93%).

show abstract

Section: Introductionmentioning

confidence: 99%

Human-guided grasp measures improve grasp robustness on physical robot

Balasubramanian

Brook

et al. 2010

2010 IEEE International Conference on Robotics and Automation

View full text Add to dashboard Cite

show abstract

“…Kang and Ikeuchi [60] proposed programming the robot by direct human demonstration. Liu et al [67] used the Barrett hand for grasping and manipulation. Kaneko and Tanie [59] presented fingers of a multifingered robot hand that touches an object.…”

Section: State Of Artmentioning

confidence: 99%

Virtual Human Hand: Grasping Strategy and Simulation

Peña Pitarch

View full text Add to dashboard Cite

La mano humana es una herramienta muy completa, capaz de adaptarse a diferentes superficies y formas, y también tocar y coger. Es una conexión directa entre el mundo exterior y el cerebro. I.Kant (filosofo alemán) definió la mano como una extensión del cerebro. En esta tesis, nosotros hemos construido una mano virtual para simular la mano humana lo más realísticamente posible. Basado en la anatomía de la mano, hemos diseñado una mano con 25 grados de libertad (DOF), con cuatro de esos grados de libertad localizados en la unión carpometacarpal, para el dedo anular y el meñique. Estos cuatro grados de libertad permiten la simulación de la mano humana cuando esta se arquea. El dedo gordo ha sido diseñado con 5 DOF, los dedos, índice y medio tienen 4 DOF, la unión metacarpofalangeal tiene dos, y las uniones interfalangeales próxima y distal tienen uno cada una. Para los dedos anular y meñique, los 4 DOF tienen las mismas uniones más los cuatro descritos arriba. El método de Denavit-Hartenberg (D-H) fue aplicado, debido a que cada dedo fue considerado como un rayo, esto es, una cadena cinemática abierta, con las uniones consideradas "revolutas". Las tablas D-H para cada dedo fueron mostradas y la aplicación de la cinemática directa e inversa permitió calcular todos los ángulos para cada unión [q1 . . . q25]T . Antes de coger cualquier objeto, nuestro sistema comprueba si el objeto esta en el espacio de la mano, mediante el análisis del espacio de trabajo. Se ha implementado un algoritmo semi-inteligente orientado a las tareas para las cuales el objeto ha sido diseñado, con el fin de tomar una decisión, una vez el usuario ha escogido el objeto y su tarea inherente. El algoritmo para coger ha sido implementado en un escenario virtual. The human hand is the most complete tool, able to adapt to different surfaces and shapes and to touch and grasp. It is a direct connection between the exterior world and the brain. I. Kant (German philosopher) defined how the hand is an extension of the brain. In this dissertation, we built a virtual human hand to simulate the human hand as realistically as possible. Based on the anatomy of the hand, we designed a hand with 25 degrees of freedom (DOF), with four of these degrees located in the carpometacarpal joint for the ring and small fingers. These four degrees permit the simulation of the human hand when it is arched. The thumb was designed with 5 DOF, the index and middle fingers have 4 DOF, in the metacarpophalangeal joint has two, and in the proximal interphalangeal joint and in the distal interphalangeal joint each have one. For the ring and small fingers, the 4 DOF are in similar joints plus as the four described above. The Denavit-Hartenberg (D-H) method was applied because each finger was considered a ray, i.e., an open chain, with joints approximated to revolute joints. The D-H tables for each finger were shown, and the application of forward and inverse kinematics permit the calculation of all angles for each joint [q1 . . . q25]T . Before grasping any object, our system checks the reachability of the object with workspace analysis. Semi-intelligent task-oriented object grasping was implemented for making a decision once the user chooses the object and the task inherent to the object. The grasping algorithm was implemented in a virtual environment.

show abstract