Search for a command to run...
This study proposes an objective method for assessing proficiency in chuck-tightening on an engine lathe using motion-capture analysis. Five experts (national certification Level 2 or higher) and five novices each performed a standardized chuck-tightening task three times under identical conditions. Full-body kinematics were recorded using fifteen markers, and clamping force was measured simultaneously. Three kinematic indices—trajectory length, velocity variability, and acceleration variability—were computed for fifteen body parts. Group differences were tested with two-sided Mann–Whitney U tests. To control for multiple testing across 15 body parts for each kinematic index, Benjamini–Hochberg false discovery rate (FDR) correction was applied (q < 0.05). Velocity variability detected the largest number of between-group differences; six of seven significant body parts remained significant after FDR adjustment. Acceleration variability was FDR-significant only at the left knee, indicating a key role in anticipatory postural control during tightening. Trajectory length showed differences at selected body parts before adjustment but did not remain significant after FDR. Clamping force was higher on average in experts (p = 0.026) and exhibited markedly lower within-subject variability, despite experts showing greater kinematic variability. These findings suggest that expert performance is characterized by controlled, task-relevant variability in movement coupled with stable force output. The proposed motion-based evaluation offers a reliable and safe complement to conventional skill assessments and can inform training and feedback systems in manufacturing education. Future work will incorporate additional kinematic features and examine effects of individual factors such as handedness, body size, and age. Trial-level reliability was high (ICC: trajectory length 0.85; velocity variability 0.82; acceleration variability 0.81).