Comparing Univariate and Multivariate Analysis of Anthropometric Measurements from 3D Body Scans for Ergonomic Work System Designs

Abstract

To design ergonomic workplaces, planners need, among other things, anthropometric data to fit the work system to the physical body dimensions of the user group. In this design process, a general distinction between univariate and multivariate approaches can be made, if several anthropometric measurements need to be considered. The aim of this publication is to present the univariate percentile approach as well as the multivariate principal component analysis (PCA) approach and to discuss differences in the resulting total accommodation (TA). A seated office workstation with visual display terminal served as a generic use case, resulting in ten relevant ISO 7250-1 measurements. The utilized anthropometric dataset, consisting of 2313 subjects (1161 men and 1152 women), was gathered between 2014-2019 within an epidemiological health study in northeast Germany, using a Vitus Smart XXL Body Scanner. With the defined use case and user group, the univariate percentile approach and the multivariate PCA approach were performed separately for the male and female subset to achieve a desired TA of 90%. In the male subset, the total accommodation was 52.7% for the univariate percentile approach and 78.3% for the multivariate PCA approach. In the female subset, the total accommodation was 51.8% for the univariate percentile approach and 78.5% for the multivariate PCA approach. Therefore, given a multidimensional use case and an anthropometric dataset in an ergonomic design process, the results of this publication indicate that it should be examined whether a multivariate approach is superior to a univariate approach to achieve an adequate TA.