AI-Based Noise Classification for Construction Sites

Effective noise monitoring on construction sites is crucial for regulatory compliance and minimizing environmental impact. This paper presents the development of an advanced noise monitoring system using artificial intelligence to classify noise events in real-time. Traditional noise measurement methods often struggle to distinguish between various noise sources, leading to inaccurate assessments and ineffective mitigation strategies. Our system leverages machine learning techniques to analyze acoustic data and identify specific construction activities, such as drilling, hammering, and heavy machinery operations. By classifying noise events, the system provides detailed insights into the major sources of noise, enabling targeted noise reduction strategies. The model was trained on a comprehensive dataset of construction-related sounds and optimized for performance in dynamic environments. Field tests demonstrate that the AI-powered system improves the accuracy of noise source identification, enhancing regulatory compliance and worker exposure assessments. The integration of this intelligent system offers a more effective approach to noise management and environmental protection. Future work will focus on refining the model's adaptability to different construction sites and expanding its application to other urban noise monitoring contexts.