Models of Human Everyday Manipulation Activities Karinne Ram´ ırez-Amaro and Michael Beetz Technische Universit¨at M¨ unchen {ramirezk, beetz}@in.tum.de Abstract. The human selection of specific postures to achieve a task among the infinity of possibilities, is the result of a long and complex pro- cess of learning. Through learning, humans seem to come to discover the properties of their bodies and the best posture to use when performing a task. This work is meanly focus on how to obtain models which describe the arm movement in common task such as the reaching movement. In other words, we are interested in extracting stereotypical motion pat- terns out of human motion observation data. The models obtained from these observation can be used to populate the knowledge base of a mobile robot in order to improve the predictive capabilities of the robot and to enable the robot to follow these stereotypical motion patterns. In gen- eral, we can observe that tasks that are executed repeatedly by a human lead to movements that are highly optimized over time, and this leads to the stereotypical and preplanned motion patterns. Is this because hu- mans look for a minimization of the muscular effort? The answer to this question is one of the main purposes of our research. 1 Introduction Unlike artificial systems, humans develop and learn how to extract and incorpo- rate new information from the environment. Animals have survived in our com- plex world by developing brains and adequate information processing strategies. Brains cannot compete with computers on tasks requiring raw computational power. However, they are extremely well-suited to deal with ill-structured prob- lems that involve a high degree of unpredictability, uncertainty, and fuzziness. Cognitive Technical Systems (CTS) are systems that are equipped with arti- ficial sensors and actuators, integrated into physical systems, and act in a phys- ical world. They differ form other technical systems in that they have cognitive capabilities including perception, reasoning, learning, and planning. These capa- bilities allow the robot to “know what they are doing ”. The cognitive capabilities will result in systems of higher reliability, flexibility, adaptivity, and better per- formance and systems. These systems are easier to interact and cooperate with humans or other systems. The importance of the study of the everyday activity research relies in the fact that it will be helpful in many ways, e.g., assist elder people or handicap people in their everyday tasks. This will increase the quality of life as well as reduce the cost of home care. Proceedings of SIMPAR 2010 Workshops Intl. Conf. on SIMULATION, MODELING and PROGRAMMING for AUTONOMOUS ROBOTS Darmstadt (Germany) November 15-16, 2010 ISBN 978-3-00-032863-3 pp. 161-170