Teaching Grasping Points Using Natural Movements

Recently, the research on robots performing every-day tasks at home, like to take care of elderly or disabled people, has pursued the problem of the manipulation of everyday objects. Among them, grasping a cloth is one of the most challenging tasks, since the textiles are highly-deformable and it is not straightforward to define a generic grasping point. In this paper, we address this problem by introducing a new robot interaction method that enables unexperienced users to control the robot in a natural way. Given a textile lying on the table the robot proposes a grasping point and the user is able to teach the robot a new grasping point. The data collected using this method is then used for training the system using a Vector Autoregression method, which produces better grasping points allowing potentially better manipulation actions. The experiments demonstrates the validity of the new interaction method and its potential to improve the point-grasping selection algorithm in different...

• Type: Journal
• Year: 2015
• Where: CCIA
• Authors: Yalim Isleyici, Guillem Alenyà
• Keywords: Artificial Intelligence, CCIA 2015