Information model for intention-based robot-human interaction

Abstract

Safe robotic-based systems gain more and more importance in an aging society. To enable all members of society to continue participation in social live, private and professional, robots are required to identify the intentions of people and the robots must be able to convey their intention to the people who use it. In this paper, an information model is introduced which matches the robot’s atomic capabilities like rotational or translational motion or grasping which must be aggregated to more complex skills. These skills must be associated with an application context. Furthermore, the mapping between the robot’s capabilities must be transparent and comprehensible to the human user, so that in fact intentions can be derived from an actual handling procedure. The information model is based on the following steps: 1st system design, 2nd gather, 3rd recognition and evaluation and 4th response.