Home /Research /Perception based dynamic trajectory generation.
PERCEPTION

Perception based dynamic trajectory generation.

Ikuo Takeuchi, Kohji Kamejima, Tomoyuki Hamada, Masahiro Tsuchiya

Year
1991
Citations
3
Access
Open access

Abstract

A dynamic motion generation method is presented for interactive robot operation. This method is aimed to make a robot execute the task of touching an object accurately with abstract instruction. The class of motion, unified regular motion and its adaptive motion, is presented within the framework of online update procedure for an a priori geometric trajectory, essentially specified in terms of task knowledge. In this presentation, planned trajectory is converted into basic input for actuators of robot with a real-time kinetics simulator and adapted through tactile and visual feedback to the object to be manipulated. The dynamic trajectory generation method is verified through experimental studies: for example, handling the object.

Keywords

TrajectoryComputer scienceTask (project management)RobotObject (grammar)Motion (physics)Computer visionArtificial intelligenceA priori and a posterioriActuator

Related papers

Browse all PERCEPTION papers