Skillful Manipulation of Electronic Musical-Note-Type Instrument Using Industrial Humanoid Robot

Abstract

Industrial robots contribute significantly to productivity improvements, quality improvements, and cost reductions at production sites. Among the industrial robots, those that can cooperate with humans without safety measures or output limitations on the motors of their axes are known as cooperative robots. In the manufacturing field, there is a growing demand for production lines where humans and robots can coexist and cooperate instead of those where robots perform all processes on their own. In this study, we focus on a cooperative robot fabricated using a 3D resin printer and aim to improve its operation using the same tools as humans. As an example of such an application, we attempt to develop a sound-feedback-based motion for manipulating an electronic musical instrument called an “otamatone”. First, the hardware for grasping the object is created using a 3D printer, and notes on the modeling process are described. We then construct an advanced sound feedback system using the Robot Operating System (ROS) to identify the sounding position and pitch of the instrument. In this study, we propose a partial model-matching method for determining the Proportional–Integral–Derivative (PID) gains of the servomotors of each joint of a robot. Consequently, the accuracy of the robot’s motion improves and the accuracy of the intended musical performance is enhanced.