Professional Documents
Culture Documents
Cognitive Systems
Robot Hands
Cognitive Systems
SS 2014 1
Lehrstuhl Informatik VI – Robotics and Embedded Systems
Fakultät für Informatik
der Technischen Universität München
Key Publications
C.L. MacKenzie and T. Iberall. The Grasping Hand. Advances in Psychology (Vol.
104), 1994
Bicchi, A.; Kumar, V., "Robotic grasping and contact: a review," Robotics and Automation,
2000. Proceedings. ICRA '00. IEEE International Conference on , vol.1, no.,
pp.348,353 vol.1, 2000
Miller, A.T.; Allen, P.K., "Graspit! A versatile simulator for robotic grasping," Robotics &
Automation Magazine, IEEE , vol.11, no.4, pp.110,122, Dec. 2004
R.M.Murray, Z.X.Li, and S.S.Sastry. A Mathematical Introduction to Robotic Manipulation.
CRC Press, 1994.
R. Haschke, J.J. Steil, I. Steuwer, H. Ritter, Task-oriented quality measures for dextrous
grasping, in: IEEE International Symposium on Computational
IntelligenceinRoboticsandAutomation,2005,pp.689–694.
A. Sahbani, S. El-Khoury, P. Bidaud, An overview of 3D object grasp synthesis
algorithms, Robotics and Autonomous Systems, Volume 60, Issue 3, March 2012,
Pages 326-336
Goals
http://www.amputee-
coalition.org/inmotion/nov_dec_07/history_prosthetics.html
Belgrade I Hand (Tomovic), ca. 1948, Belgrade/USC II Hand (Tomovic/ Stanford-JPL hand (Ken Salisbury),
1 DOF, flexing fingers Bekey), ca. 1988, 4 DOF, flexing fingers, ca. 1985, 2 flexing fingers, one
rotating thumb thumb
Situation Today
• Today, there is a growing interest in
precision manipulation across all fields
of robotics
• There is also continued interest from
the military, see the DARPA program on
Revolutionizing Robotics
• Another activity is Exoskeletons for
enhancing a human‘s (a soldier‘s)
power and stamina
Johns Hopkins University's Bimanual
Dexterous Robotics Platform
http://www.gizmag.com/darpa-advanced-prosthetics-bomb-disposal-robot/27975
Goal:
To hold an object in the robot’s end-effector firmly, preventing loss of contact. To maintain
the grasp even in the face of unknown disturbing forces or moments applied to the object.
Definitions
Grasping
• Holding something firmly relative to the end-effector
Form closure
• The object is completely enclosed by the hand, without
any wiggle room. No movement is possible in any direction
(assuming that all contacting bodies are rigid)
• Seven contact points are necessary for an arbitrary object in 3D with 6DOF [Bicchi,
Antonio. "On the closure properties of robotic grasping." The International Journal of
Robotics Research 14.4 (1995): 319-334.]
Force closure
• The object is held applying some force.
• “Similar to form closure, but relaxed to allow friction forces to balance the object
wrench” [Handbook of Robotics,Siciliano, Khatib (Eds.), Springer 2008, p684]
• 2 or 3 contact points necessary for a 3D object with 6DOF,
SS 2014depending on the type of contact
Cognitive Systems 9
Lehrstuhl Informatik VI – Robotics and Embedded Systems
Fakultät für Informatik
der Technischen Universität München
Underactuated grippers
FESTO Fingripper
Anthropomorphic hands
Soft hands
http://www.robotics.tu-berlin.de/fileadmin/fg170/Publikationen_pdf/deimel-14-RSS-draft.pdf
Hand models
http://www.youtube.com/watch? v=0d4f8fEysf8
http://grasp.xief.net/documents/taxonomy.pdf
http://grasp.xief.net/documents/taxonomy.pdf
http://grasp.xief.net/documents/taxonomy.pdf
http://grasp.xief.net/documents/taxonomy.pdf
http://grasp.xief.net/documents/taxonomy.pdf
Assumptions:
• hand links are rigid
• the object is rigid
• each contact point has a unique, well-
defined tangent plane
Object Modeling
Stereo cameras
Time-of-flight
cameras
BioTac sensor:
- force
Strain gauges to
- vibration
measure forces
- temperature
RGB-D sensors
Arms
The logical clauses A, B and C defining the transition conditions, are the following:
A = no−object, B = no−grasp and C = collision ∨ failed-grasp
Chavdar Papazov, Sami Haddadin, Sven Parusel, Kai Krieger, and Darius
Burschka. Rigid 3D Geometry Matching for Grasping of Known Objects in
Cluttered Scenes. International Journal of Robotics Research, 31, April 2012
Beyond grasping
Beyond grasping
• Grasping is usually limited to holding the objects in the end-effector.
• Most robots are limited to pick-and-place tasks
• In-hand-manipulation requires much better dexterity
In-hand manipulation:
• Taking an object, and repositioning it in the hand
• Rotate and move the object to ease hand-over
• Or even grab a device and touch buttons
• Using an object effectively as a tool
In-hand manipulation
http://www.k2.t.u-tokyo.ac.jp/members/senoo/paper/senoo_icra09.pdf
http://spectrum.ieee.org/automaton/robotics/humanoids/dlr-super-robust-robot-hand
Anthropomorphic hands