Investigating the Impact of Construction Noise on Work Posture and Productivity with Virtual Reality–Based Masonry Simulation

Abstract Construction sites are often characterized by high-intensity noise, which routinely exposes workers to considerable acoustic loads generated by machinery and heavy equipment. Excessive noise contributes to auditory damage and interferes with communication, cognitive processing, and physical control. However, the influence of workplace noise on workers’ physical behavior and its subsequent effects on ergonomic risk and productivity remains unexplored. This study experimentally investigated how different levels of construction noise affect ergonomic responses in terms of biomechanical stress and task performance. To this end, controlled experiments were conducted in a virtual reality (VR) environment, which allowed for precise manipulation of noise levels while providing safe, immersive, and reproducible simulation of construction tasks under varying noise conditions. In the experiment, participants performed lifting and placing motions during masonry tasks under noise levels ranging from 50 to 100 dB in 10-dB intervals. The motions were captured using vision-based three-dimensional (3D) pose estimation and then analyzed using biomechanical simulation to estimate joint moments and spinal compression forces. Quantitative biomechanical assessments were conducted on key body regions, including the lower back, torso, and shoulders. The experiment results indicated a notable increase in compressive forces at the lower back (5.74% at L5-S1 and 7.08% at L4-L5) as noise intensity increased, and variations in torso movement were associated with declines in productivity. The results suggest that noise exposure was associated with posture-related biomechanical indicators and productivity in a body-part-specific manner, with clearer patterns for the lower back and torso than for several other posture-related measures. Accordingly, this study provides quantitative evidence on noise-intensity-related, body-part-specific biomechanical responses and task performance in a controlled VR masonry task, which may inform noise-control and ergonomic planning efforts.