Identifying Alternative Movement Techniques from Existing Motion Data: An Empirical Performance Evaluation

A manual task can be performed based on alternative movement techniques. Ergonomic human motion simulation requires consideration of alternative movement techniques, because they could bring different biomechanical, physiological, and psychophysical consequences. A method for identifying movement techniques from existing motion data was developed. The method is based on a JCV (Joint Contribution Vector) index and statistical clustering. A JCV quantifies a motion's underlying movement technique by computing contributions of individual body joint DOFs (degree-of-freedom) to the achievement of the task goal. Given a set of motions (motion capture data) achieving the same or similar task goals, alternative movement techniques can be identified by 1) representing the motions in terms of JCV and 2) performing a statistical clustering analysis. Performance of this movement technique identification method was evaluated based on a set of stoop and squat lifting motions. It was found that the method was able to identify the two distinct lifting techniques from the lifting motion dataset. Combined with the motion modification/adapting/editing methods, the movement technique identification method will enable consideration of alternative movement techniques in human motion simulation.