|Hands.dvi ECHORD Project
Contents 1 What is HANDS.DVI?
2 Partner's Meeting
3 SynGrasp Toolbox
1 What is HANDS.DVI?
Hands.dvi is an experiment project under the European project "European Clearing House for Open Robotics Development" ECHORD
, Grant Agreement n. 231143.
The scientific goal of HANDS.DVI (A DeVice Indipendent programming and control framework for robotic HANDS) consists of developing a common framework to programming robotic hands independently from their kinematics, mechanical construction, and sensor equipment complexity. Recent results on the organization of the human hand in grasping and manipulation are the inspiration for this project. The human hand is a redundant structure, with many degrees of freedom with respect to the dimension of the controlled workspace. The force control strategies we use for grasping and manipulating objects depends on our sensors and actuators. The hand sensorimotor transformations impose constraints between the degrees of freedom, so that the number of effective degrees of freedom is much lower than the number of joints. The reduced set of parameters that we effectively use to control our hands is known in the literature as the set of synergies.
The synergistic organization of the human hand is the theoretical foundation of the innovative approach to design a unified framework for robotic hands control.
We will develop a kinematic model of an anthropomorphic robotic hand that embodies synergies. We will refer to this model as the paradigmatic hand which will be used to design a new unified synergy language to control robotic hands. Theoretical tools will be studied to design a suitable mapping function of the control action (decomposed in its elemental action, synergies) from the paradigmatic hand domain onto the articulated hand co-domain. The definition of this mapping is the core of HANDS.DVI. It will consists of elemental actions or synergies engaged in a sequence whose dimension, or complexity, increases with the complexity of the manipulation task to be performed.
The advantages and the limits of our approach will be assessed both analytically and with numerical simulations. Finally, the developed control framework will be applied on an experimental set up consisting of three robotic hands with dissimilar kinematics grasping an object instrumented with force sensors.
The ultimate goal of the project is to design an innovative architecture for the control of robotic hands, whose main advantage with respect to the state of the art is the significant reduction in the number of controlled degrees of freedom. Furthermore, at the synergy level the control will be device independent, leading to a sort of standardization between control strategies among robotic hands with dissimilar structures.
These results will encourage the diffusion of robotic hands in advanced industrial environments, such as hyper flexible cells. Furthermore, the design of robotic control systems that think as human hands, will ultimately promote human-robot interaction and cooperation
2 Partner's Meeting
IIT, Genova 24th January 2011
This has been the first formal meeting between all the partners.
It was a nice occasion to share information and ideas about the project.
In the follow you can find some of the presentation given during the meeting.
Presentation of the project
- Prof. Domenico Prattichizzo
Review of the literature on mapping techniques
- Guido Gioioso
Review of the literature on eigengrasps techniques
- Gionata Salvietti
Presentation of the activities at University of Pisa
- Alessandro Ossino
3 SynGrasp Toolbox
This is the preliminary version of the Matlab toolbox SynGrasp (Synergy
Grasping). Its main features are summarized in the following:
- Hand model definition: it is possible to define new hand models
in addiction to those already included in the toolbox (the human
paradigmatic hand, the Barrett hand and the DLR-I
Hand). Denavitt-Hartenberg parameters have been chosen as default
notation to define a new hand model. A synergy matrix can be
associated to each hand. The matrix can be manually set or dynamically calculated.
- Object definition: virtual objects can be created for grasping
simulations in the virtual environment. The user can define their
shape and position with respect to the hand. Some simple objects are
available at the moment.
- Grasp definition and analysis: the grasping hand configuration,
contact points, grasp and Jacobian matrix can be defined in order to
evaluate grasping performances. Controllable internal force
subspace, rigid-body object motion, manipulability ellipsoids can be evaluated.
- Grasp force optimization: the set of internal forces that allows
minimizing suitable cost functions can be computed.
The first realise of our Matlab Toolbox can be downloaded here
IROS 2011, Presentation of HANDS.DVI at the Workshop European Efforts in Strengthening the Academia-Industry Collaboration [Proceedings]
AUTOMATICA.IT 2011, Presentation of the HANDS.DVI experiment,[Poster]
On-line our summaries on the Echord website [link]
ECHORD at the European Robotics Forum 2011: presentation of the Hands.dvi Project. [slides]
ICRA 2011, Workshop on Manipulation Under Uncertainty: presentation of the poster Mapping Synergies from Human Hand to Robotic Hands with Dissimilar Kinematics: an Object Based Approach
File translated from
On 4 Oct 2011, 11:46.