Human Driven Robot Grasping: An Interactive Framework
One main problem in the field of robotic grasping is to teach a robot how to grasp a particular object; in fact, this depends not only on the object geometry, but also on the end-effector properties. Different methods to generate grasp trajectories (way-points made by end-effector positions and its joint values) have been investigated such as kinaesthetic teaching, grasp recording using motion capture systems, and others. Although these method could potentially lead to a good trajectory, usually they are only able to give a good initial guess for a successful grasp: in fact, obtained trajectories seldom transfer well to the robot without further processing. In this work, we propose a ROS/Gazebo based interactive framework to create and modify grasping trajectories for different robotic end-effectors. This tool allows to shape the various way-points of a considered trajectory, and test it in a simulated environment, leading to a trial-and-error procedure and eventually to the real hardware application.
KeywordsObject grasping and manipulation Grasp synthesis Grasp planning Grasp simulation Human in the loop
- 2.Quigley, M., Conley, K., Gerkey, B., Faust, J., Foote, T., Leibs, J., Wheeler, R., Ng, A.Y.: ROS: an open-source robot operating system. In: ICRA Workshop on Open Source Software, vol. 3, p. 5 (2009)Google Scholar
- 3.Rviz - 3d visualization tool for ROS. http://wiki.ros.org/rviz
- 4.Koenig, N., Howard, A.: Design and use paradigms for Gazebo, an open-source multi-robot simulator. In: Proceedings of 2004 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2004), vol. 3, pp. 2149–2154. IEEE (2004)Google Scholar
- 9.Santaera, G., Luberto, E., Serio, A., Gabiccini, M., Bicchi, A.: Low-cost, fast and accurate reconstruction of robotic and human postures via IMU measurements. In: 2015 IEEE International Conference on Robotics and Automation (ICRA). IEEE (2015)Google Scholar