Abstract
Task simulation and workplace evaluation under the digital environment have been widely adopted at early production planning stages by large-scale manufacturing companies. In order to successfully simulate the manual operation, digital human model (DHM) was created as a simulation tool to help managements making better decision. However, the unpredictability and complexity of human movements increase the difficulty of creating an accurate and efficient simulation model for human movements. Herein, a feasible approach was proposed to enhance the simulation reliability by retrieving the working posture parameters from collected human motion data and inputting to the digital human motion generation system. The main purpose of this study was trying to gather critical information from workers in an assembly line by the means of the motion capture technology. An eight-subject experiment was designed to measure the pelvis orientation and position when approaching a working spot from five initial positions, and analyze the working posture while pointing a target, grasping a know, and pneumatic wrenching at six working heights.The results indicated that the position and orientation of the subject’s whole body (while performing the operation) were significantly affected by his or her approach direction. Linear relationships between the approach direction and the final whole-body orientation were also obtained for each of the three tasks. Based on the result, we inferred that the subjects would prioritize their work efficiency and would try to find a comfortable way to perform their jobs. They deviated from the least-distance (straight line) walking path and reoriented their whole bodies for the operation when approaching the working spot. In addition, the working postures for the three tasks at six working heights were analyzed, too. The subjects tended to adopt standing postures when working at a height between eyes and hip, and stooping or squatting postures when operating at a height between knee and ankle. The findings of this research can be used to enhance the digital human modeling motion generated for human movement simulation.