Impact Task Model and Acoustic Recognition System With a Humanoid Robot for an Automation of Hammering Test

The purpose of this study is to automate the hammering tests by using a humanoid robot. Therefore, this report discusses the impact of the task model using the robot and a new acoustic recognition system. This study also deals with the study of the robot playing the glockenspiel. We proposed a flexible rubber stick for the humanoid robot in order to control the impact task. The robot hits the glockenspiel using the vibrations of a flexible rubber stick. The results showed that the robot achieved high accuracy impact with the stick. Also, the robot could move both arms efficiently. Regarding the acoustic recognition, we focused on the frequency and the sound pressure generated from objects. The results showed that the acoustic recognition system detected the sounds. This system also recognized the sound of every object. Moreover, on the basis of the impact task model and acoustic recognition system, we developed a hit sound feedback system. This system achieved efficient results in the case study. Therefore, the proposed approach demonstrated the efficiency of using a humanoid robot in an automated hammering test.