Diplomarbeiten, Studienarbeiten, Praktika
Internship or Master Thesis: Autonomous Grasp Planning with Robotic Hands
6. September 2011
Object grasping and manipulation has become an area of great interest in robotics, especially due to the development of dexterous grasping devices like robot hands that increase the flexibility and versatility of the robot arms. Despite great advances in the mechatronic development of robot hands, there are still large gaps in the development of autonomous grasp planners.
Two types of grasps are mainly considered for multi-fingered hands: power and precision grasps. Power (enveloping) grasps use the whole surface of the inner hand to restrain the object, and precision (fingertip) grasps use only the fingertips to grasp the object. In fact, the two types of grasps are easily identified in the grasp actions performed by the human being, where fingertips and distal phalanges are used for fine manipulation, and the inner parts of the hand (palm and proximal phalanges) are used to firmly restrain an object. Most of the theoretical work on grasp synthesis is focused on computing precision grasps, as they are obtained with few fingertip contacts (3-5). Power grasps are mainly determined by the geometrical characteristics of the hand and object.
An autonomous grasp planner is currently under development. The work of the student will be the evaluation of the current grasp strategies, and the creation and implementation of new strategies for both power and precision grasp planning. The algorithms should be initially validated using the virtual environment, and later implemented on a real arm-hand system.
|Fig.1 Precision and power grasp with the DLR 4-finger hand.|
- Good knowledge of Python and C++
- Knowledge of OpenRave is a plus.
- Knowledge of Matlab/Simulink is desired.
Please feel free to contact me should you be interested and/or need additional details.