Abstract
The objective of this study was to automate the dual-arm coordination of a humanoid robot when using a musical saw as a percussion instrument to produce a target sound. We examined the striking motion of the robot using its elastic stiffness, the change in vibration caused by the striking tool resulting from the robot arm-applied force on the saw, and dual arm cooperative motion based on striking sound feedback. To accomplish this, the frequency characteristics of the musical saw and the location of sound generation were determined. Based on the characteristics of the striking task and using an acoustic recognition system, we then developed an impact sound feedback system that could provide performance control of the two humanoid robot arms. The results indicate that humanoid robot can play the musical saw by controlling the curvature freely to achieve target sounds.
Local Environment Recognition System Using Modified SURF-Based 3D Panoramic Environment Map for Obstacle Avoidance of a Humanoid Robot
