UPPER-BODY REACH POSTURE PREDICTION FOR ERGONOMIC EVALUATION MODELS

Proper assessment of human reach posture is one of the essential funct ions for ergonomic workspace design and evaluation in computer-aided e rgonomic evaluation models or any CAD system with a built-in man model . To predict reach posture, most existing models have been using heuri stic methods, which provide only the range of feasible postures, not a lways ensuring the one that a person naturally takes. An analytic reac h prediction algorithm was developed in this study by employing the in verse kinematics methods. Each upper limb is modelled as a four-link s ystem, consisting of trunk, upper arm, lower arm, and hand, being rega rded as a redundant manipulator with a total of eight degrees of freed om. Among several kinematic methods for solving human reach movement, the resolved motion method which is one of the redundant manipulator t echniques in robotics was found to be effective. In this method, the j oint range availability was used as a performance function to guarante e kinematic optimality and to simulate human reach closely. In additio n, an approximate algorithm to generate the workspaces of human body w as developed. Real reach postures taken from the subjects were analyze d by the motion analysis system and were statistically similar to thos e obtained from the prediction model.