Earticle

Inappropriate work postures in industry can cause musculoskeletal fatigue and strain, leading to work inefficiency, quality problems, and safety accidents. Therefore, a system capable of recognizing and intervening in dangerous postures during work in real time is needed not only for industrial safety but also for project performance management. This study proposes a 2-channel real-time work posture management system that combines RULA analysis based on frontal upper body images and REBA analysis based on lateral full-body images. The proposed prototype was implemented using synchronized frontal and lateral cameras and was preliminarily applied to three participants performing terminal connection tasks in an actual manufacturing environment under no-alarm and alarm conditions. The system records posture risks every second and generates an alarm sound if a dangerous posture persists in the same body part for more than 3 seconds. As a result of the preliminary application, the rate of adopting dangerous neck postures under alarm conditions decreased. The maximum time maintained in dangerous neck postures was also reduced. However, since this study was conducted with only three participants and a single task condition, these results should be interpreted as preliminary results demonstrating the applicability of the prototype rather than as general effects.