Explanation-based Manipulator Learning: Acquisition of Assembling Technique through Observation

This paper describes a robot manipulator system currently under development which learns assembling technique from observation of skilled worker’s demonstrations. The system is an application of EBL (Explanation-Based Learning) to the robot manufacturing domain. The system improves the operating capability of the robot manipulator through observing, analyzing the skilled worker’s assembling operation, and learning the assembling technique, that is, the tacit knowledge of skilled workers which makes them complete the assembling task more efficiently. The learning process is a knowledge-based deduction approach, requiring sufficient background knowledge, to understand the observed sequence.