Identification of Modular Construction Activity Noise Levels by using K-Means Clustering

In recent years, modular construction manufacturing is increasingly applied as a construction method due to its cost-effectiveness and shorter construction time. However, the assessment of noise exposure to workers in the off-site construction facilities has not been fully studied yet. This paper investigates a framework which assesses the noise exposure using a probability-based prediction approach. The framework consists of: (1) collecting procedure information of manufacturing production lines; (2) analyzing task duration probability statistics; (3) measuring the task-based noise levels; (4) separating the noise level of each activity by using clustering analysis. The results of this study will be beneficial by providing precise uncertainty of the noise measurement and possible noise control recommendations.