The research results signify a significant leap in automating waste monitoring within construction and demolition settings. By utilizing the YOLO v8 Nano model, the study accomplishes real-time waste detection, markedly improving operational efficiency and generating substantial cost savings. This real-time detection ensures the prompt removal of waste, reducing the occurrence of overfilled dumpsters and the accompanying operational expenses. Furthermore, the research introduces an innovative aspect by establishing a database of waste dumping events. This database allows for continuous progress monitoring, offering valuable data insights into waste disposal patterns, quantities, and types. Consequently, this data serves as a crucial foundation for data-driven decision-making and the refinement of future waste management strategies.
Moreover, the research introduces a custom dataset meticulously tailored to the specific demands of waste monitoring. This dataset encompasses three distinct classes: Vehicle, Human, and Dumpster. It functions as a critical resource for training and evaluating the YOLO v8 Nano model, ensuring its precision and suitability for real-world applications. This dataset not only supports the research’s objectives but also contributes to the broader academic community as a benchmark for future research in object detection and surveillance systems. Expected results from this research extend to both academic and practical applications, enriching the fields of computer vision and deep learning and facilitating advancements in automated waste monitoring practices.
The research is expected to make significant contributions to both academic and industry domains. In academia, it will show the potential of Computer Vision in the construction field by showcasing the practical application of object detection for real-time waste monitoring. The research outcomes will serve as a valuable case study for future studies in object detection and automated surveillance systems. In the industry, the research offers practical solutions to a long-standing problem in construction and demolition waste management. It introduces a cost-effective and efficient approach that construction companies, waste management firms, and regulatory bodies can adopt. Automating waste monitoring and creating a database of waste dumping events will lead to tangible benefits such as reduced operational costs, streamlined waste management, and improved sustainability practices, thereby contributing to a more environmentally responsible construction industry.
