The manipulation part aims at developing technologies for implementation of human-like manipulation skills by task skill learning and control methods. This part majors three techniques: (a) skill learning by imitation that is to learn complex and precise manipulation skills like humans, (b) autonomous primitive skill learning that is to find how many basis skills exist given task, and to model them autonomously, (c) Grammaticalization that is to reuse basis skills variously, as words in linguistics.
Skill learning by imitation - Techniques to learn a complex and precise manipulation skill like a human
Autonomous primitive skill learning - Techniques to find how many basis skills are in the given task, and model them autonomously
Grammaticalization - Techniques to reuse the basis skills variously, as the words did in Linguistics
2010.10.22~2011.08.21 Development of a biology-inspired learning architecture and task skill learning technology KIST