An Efficient Method for Modeling an Individual’s Perception of Postural Stress

Abstract

Various Posture Analysis Tools (PAT), for example, OWAS and RULA, are currently used to evaluate and control work-related postural stresses. These tools do not consider substantial individual differences in the perception of postural stresses. Consequently, they do not provide an insight into the population stress distributions of different working postures. The limitations make it difficult to provide protections in individual-specific manners and accurately estimate the effects of job design at the population level. An alternative to the existing PAT is to develop individual-specific PAT and use them for protecting individuals. With a large database of such individual-specific PAT, population level analyses also become possible. As an initial effort towards individual-specific and population-level posture analyses, our goal is to develop a procedure for creating individual-specific PAT. Our hypothesis is that individuals’ specific PAT can be represented as simple mathematical functions, referred to as posture-stress mapping (PSM) functions. To test the feasibility of the PSM function concept, a pilot study was conducted. As for the common basic function form, a weight sum of body joint moments was tested. Six subjects participated. Each subject performed a 20-second static posture holding trial for a set of 180 working postures and conducted a subjective rating of postural stress using the Borg CR10 scale. The six subjects differed from one another significantly in the distribution of the measured postural stresses justifying the needs for the individual-specific modeling approach. For each subject, the posture-stress dataset was divided to form a model test dataset and differently-sized model building datasets. Each of the model building dataset was fitted into the base function form. The resulting PSM function was used to predict the postural stress responses for the postures included in the model test dataset. Its prediction accuracy was evaluated by comparing the actual and predicted postural stresses. For two of the six subjects, their PSM functions were found to be able to accurately predict postural stresses of working postures. Overall, the current results suggest that individual-specific modeling of postural stresses seems feasible; however, different base function forms will need to be evaluated in the future.