Avatar-based human posture analysis and workplace design

Abstract

Stressful working postures are associated with increased risks of work-related musculoskeletal disorders (WMSD). The reduction of stressful working postures is necessary for the safety and health of workers and is also related to increased productivity and worker satisfaction. The evaluation and control of work-related postural stress is currently achieved through a process that requires observation, analysis, and redesign (OAR approach). As an alternative overcoming the limitations of this approach, the authors have previously proposed an APAD approach [1]. The objective of the current paper is to report the initial results of the APAD research focusing on the modelling of individual PSM functions. 3 participants were asked to perform short-period static posture holding for 108 combinations of posture and hand load weights. Immediately after each task trial, the participants performed subjective ratings using the Borg CR10 scale to quantify the level of perceived postural stress. Individual PSM functions were derived through regression, and were evaluated by calculating the average absolute prediction error. The study results may serve as a basis for creating a large set of avatars that perceive postural stresses in a way similar to their actual human counterparts. Such avatars could be used to simulate various work situations in the digital world to aid in the design of less stressful and ergonomic workstations and work tasks.