2012 Volume 2012 Issue KST-15 Pages 05-
This paper presents a method to evaluate a person's skill level for metal filing. We used wearable accelerometer/gyroscope sensors and collected data from 4 expert coaches and 10 learners. The data are analyzed from the viewpoint of the mechanical structure of their bodies during metal filing. Our analysis yielded two effective measures for skill assessment: "Class 2 Lever-like Movement Measure" and "Upper Body Rigidity Measure". The weighted total measure succeeded in distinguishing the coach group and the learner group as individual skill level groups at a 95% confidence level. The highest-level learner, the lowest-level learner, and the group of other learners were also able to be distinguished as individual skill level groups at a 95% confidence level; this is the same result as an expert coach's subjective score.