Development and Evaluation of Regression Models for the Prediction of Human Torque Strength Applied in the Prone Position

Abstract

This paper describes an experimental investigation conducted to develop and test predictive models for human torque strength in simulated maintenance tasks performed in the prone position. Eight linear regression models were developed for the prediction of human torque strength applied with a socket wrench while lying in the prone position. The models utilized anthropometric and strength data from 40 subjects in conjunction with task point location variables in their formulation. The predictability of each model was evaluated by comparing the observed and predicted torque values from 10 additional subjects.