Object Pre-Grasp Configuration of a Robot Articulated Hand in a Virtual Environment

Autor: A. Bouachari, Jean-Francois Brethe, A. Haddad
Rok vydání: 2018
Předmět:
Zdroj: 2018 International Conference on Signal, Image, Vision and their Applications (SIVA).
DOI: 10.1109/siva.2018.8660970
Popis: The grasp of an object from a specific location, its handling and delivery in a well-defined working environment require knowledge concerning both manipulators and mobile robots. The positioning of a robot arm with the object involved in the mobile manipulation requires the calculation of the pre-grasp configurations as well as that of the selected final grasp that determines the final position of the robotized hand relatively to that of the object. The present research uses the 'gazebo-ros' simulator to locate the position of the hand relatively to that of the object. This is applied to the case of the ‘care-bot-3’ service robot where a global navigation software package is used to animate-simulate the initial functionality of the mobile robot in order to bring the object under consideration in the working environment of its arm. The robot begins its approach in local navigation form, the positioning in the horizontal plane of the object taking place on the basis of the pre-capture configurations related to the industrial manipulator '70DF'. The generation of the capture configurations is carried out through an inverse kinematic model developed under ‘ROS’ environment on its final articulation ‘arm_7_link’ in its basic referential ‘base_link’. The displacement of the object to its final position is carried out through a data set of trajectories of the object defined between the primitives of the hand 'schunk' mounted on the manipulator of ‘care_bot’. The efficiency of the approach developed is demonstrated through its application to the virtual environment 'gazebo' as well as on the display-visualizer ‘RVIZ’ via a kinect camera.
Databáze: OpenAIRE