Open AccessAnalysis of Wrist Hand Motion for Monitoring of Basic Welder Training using Wearable Sensors
Author(s) -
Triwilaswandio Wuruk Pribadi,
Takeshi Shinoda
Publication year - 2022
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/972/1/012010
Subject(s) - accelerometer , wearable computer , wrist , motion (physics) , computer science , match moving , computer vision , artificial intelligence , welding , gyroscope , simulation , engineering , mechanical engineering , medicine , embedded system , surgery , aerospace engineering , operating system
During the training of a welder, either novice or professional, most activities are focused on the acquisition of wrist-hand motion skills. In the basic welding training, trainees initially required hand-on practices to acquire the skills of wrist hand motion to maintain the distance of electrode tip to a base metal such that the welding arc was continuously flaming. Secondly, trainees were practices of manipulating hand motion to follow seam tracking for joining two metals within defined speed & torch height. These practices were then continued for various types of weld joints. The result of acquiring this skill level was then assessed by inspecting the visual appearance of the weldment. In this study, an effort was undertaken to monitor and assess the progress of acquiring wrist-hand motion skills using wearable sensors: accelerometer, gyroscope, and magnetometer. Then, the record of those sensors was plotted as a time series signal compared with those performed by the training instructor. Their achievement of skills grade was analyzed using the Supervised Vector Machine (SVM) Learning Method. The result has indicated that this proposed method can assist in assessing welder trainees’ efforts to improve their skills.