第59回創成塾

[2013.02.13]


2013年2月19日(火) 18:00-20:00 豊中キャンパス 基礎工学研究科 A403にて第59回創成塾を開催致します.

内容:
18:00-18:30 Salvatore Anzalone(post-doc researcher, the Institute for Intelligent Systems and Robotics from Universitè Pierre et Marie Curie
18:30-19:00 Q&A and discussion with students
19:00-19:30 Fu Kin Chung Denny (D3, Graduate School of Engineering and Science, Osaka University)
19:30-20:00 Q&A and discussion with students

基礎工学研究科 A403講義室

講演1:Salvatore Anzalone(post-doc researcher, the Institute for Intelligent Systems and Robotics from Universitè Pierre et Marie Curie)
” Towards personal robotics.”

Robotic researchers focused their works on making robots completely autonomous in real environments. According to this approach, robots are something like a tool or a sort of “intelligent device” to be employed to achieve missions instead of humans. This implies also a master-slave relationship between the human and the robot. However, robots can collaborate with humans in a more strict way being a social, cooperative and communicative partner. But humans consider themselves individuals with an unique personality, so they expect to be treated likewise. To be perceived as active and effective partners, truly accepted by humans, social robots need to learn how to relate with the individuality of humans: they need to recognize them and adapt their behaviors, vocabulary, and social rules, according to the individuals that are involved in the interaction, their personality, their personal preferences, and their social role. After a short introduction to Social Robots, in this talk some experiences about multimodal (video and audio) people identification for robotic agents will be shown. Then a robot able to profile human partners by the learning of their preferences will be presented and results will be exposed.

講演2:Fu Kin Chung Denny (D3, Graduate School of Engineering and Science, Osaka University)
” A studies about motor synergies in optimal control of musculoskeletal robots.”

The studies in the field of human motor control are about how we control our bodies. It can be imagined that the control is very difficult because of the high degree-of-freedom and redundant characteristics of the musculoskeletal structure. Motor synergies have been proposed as a control solution to tackle the degree-of-freedom and redundant characteristics. Among many interpretations of motor synergies, one suggests that many muscles are not controlled individually. Instead fewer groups of muscles with specific activation patterns, namely motor synergies, are coordinated. Control can then be simplified by coordinating fewer control variables in terms of motor synergies. Moreover, it has been shown that the optimal control theory from engineering can be used as a computational theory which can predict human-like motion, but the optimal control theory also suffers from so called the curse of dimensionality when the system dimensionality (degree-of-freedom) is large. Our research is about how the concept of motor synergies from the human motor control can be applied to facilitate the solving of engineering optimal control problem. In this talk I will present simulation results using different types of motor synergies. From the results we will see how the control complexity can be reduced, and also some hints about how to obtain motor synergies for achieving novel tasks.

[ページ先頭へ]