Assessing the Accuracy of Ergonomic Analyses when Human Anthropometry is Scaled in a Virtual Environment 2006-01-2319

This study addressed the effect of scaling subjects in a virtual reality (VR) environment when performing ergonomic evaluations for assembly automotive tasks. Subjects were selected to fit into one of 4 anthropometric groups, ranging in size from a 5^th percentile female to a 95^th percentile male. Each subject performed 3 tasks while interacting with a digital rendering of a vehicle. Subjects were represented as a human manikin (Classic Jack 4.0, UGS) whose actions were driven by motion tracking (EvaRT, MotionAnalysis). Each subject performed the tasks under 4 different conditions; once with unscaled anthropometry and in three conditions where they were made to appear either larger or smaller than their actual height. Peak and cumulative low back loads and shoulder moments, as well as joint angles, were calculated and compared (ANOVA). In some cases subject scaled to a particular size performed differently than those that were actually that size. Guidelines have been provided for selecting subjects for ergonomic studies performed with motion capture and virtual reality integration.