Ergonomics modelling and evaluation of automobile seat comfort

Abstract

Automobile seats are developed in an iterative manner because subjective feedback, which is usually of questionable quality, drives the design. The time and cost associated with iteration could be justified if the process was guaranteed to produce a comfortable seat. Unfortunately, this is not the case. Current practices are based on the premise that seat system design teams need objective, measurable laboratory standards, which can be linked to subjective perceptions of comfort. Only in this way can predictions be made regarding whether or not a particular design will be viewed by the consumer as comfortable. This type of forecasting ability would effectively improve the efficiency with which automobile seats are designed. In this context, the research reported, developed, and validated a stepwise, multiple linear regression model relating seat interface pressure characteristics, occupant anthropometry, occupant demographics, and perceptions of seat appearance to an overall, subjective comfort index derived from a survey with proven levels of reliability and validity. The model performance statistics were: adjusted r(2)=0.668, standard error of estimate=2.308, F (6, 38)=15.728, p=0.000, and cross-validated r (15)=0.952, p=0.000. From the model, human criteria for seat interface pressure measures were established. These findings could not have been attained without first demonstrating that (1) the data collection protocol for seat interface pressure measurement was repeatable and (2) seat interface pressure measurements can be used to distinguish between seats.

A conceptual framework proposed to formalize the scientific investigation of automobile seat comfort

“…It is essential that researchers formalize the implied relationship by developing predictive instruments linking performance measures (obtained through scientifically sound and standardized methodologies) to subjective perceptions of comfort (derived from a questionnaire with proven levels of reliability and validity). Modeling alternatives include statistical techniques (Kolich and Taboun, 2004) and artificial neural networks (Kolich, 2004). The models may ARTICLE IN PRESS also be improved by considering the previously described social factors.…”

Analysis of sitting forces on stationary chairs for daily activities

“…Most sitting force studies have been focused mainly on static sitting for sitters weighing less than 136 kg, in wheelchair research (Gilsdorf et al 1990;Gilsdorf et al 1991;Aissaoui et al 2001;Chou et al 2002) and vehicle seat design (Rakheja et al 2002;Kolich 2004;Kolich and Taboun 2004;Tarczay 2005;Bush and Hubbard 2007;Moes 2007). Sitting force studies were also found in sitting comfort research with emphasis on sitting pressure distribution on a seat (Chow and Odell 1978;Drummond et al 1982;Bush and Hubbard 2007), which data can be converted to sitting forces.…”

Identification of sensory variables towards the integration of user requirements into preliminary design

“…That late sensory validation is part of the trial-and-error method. Kolich and Taboun (2004) illustrate this iterative manner on the automobile seats design case. Integrating user requirements into preliminary design ensures less iterations in the design process.…”