Real-Time PPE Monitoring And Demographic Analysis At Construction Sites Using Yolo
Abstract
The construction industry poses significant challenges in maintaining worker safety and rights. Existing safety measures involve an on-site officials working on the ground to ensure that worker safety measures like wearing safety helmet and power monitoring are undertaken as mandated. However, these methods are often inefficient, as on-ground checking takes significantly more time and allows the construction site authority to plan their excuse. In response, this paper presents a comprehensive novel approach, an AI-enhanced construction site monitoring system that integrates real-time CCTV hazard detection, age verification, and personal protective equipment (PPE) compliance as robust measures to ensure worker safety. Leveraging a React Native application for worker-side interaction and officer-side oversight and integrating real-time location tracking and Firebase-based notifications, the system fosters proactive safety measures and promotes worker rights. Compared to existing solutions, the proposed system boasts superiority with an average accuracy of 92.5%, scalability with the use of Android-based applications and modifiable APIs and AI models, and user-centric features, leading to a safer and more transparent system that surpasses existing solutions.
References
[2] D. Ren, T. Sun, C. Yu, and C. Zhou, “Safety Helmet Detection for Construction Site: A Research Study”, in Proceedings of the 2021 International Conference on Computer Information Science and Arti- ficial Intelligence (CISAI), Kunming, China, 2021, pp. 186-189. DOI: 10.1109/CISAI54367.2021.00042.
[3] W. Xiaoyan, C. Pei, L. Rui, Y. Bo, and N. Wenzhao, “Intelligent Analysis Technology of Power Monitoring Video Data Based on Convolutional Neural Network: A Research Perspective”, in Proceedings of the 2020 5th International Conference on Mechanical, Control, and Computer Engineering (ICMCCE), Harbin, China, 2020, pp. 2461-2464. DOI: 10.1109/ICMCCE51767.2020.00531.
[4] F. Alsakka, S. Assaf, I. El-Chami, and M. Al-Hussein, “Computer Vision Applications in Offsite Construction”, Automation in Construction, vol. 154, p. 104803, 2023. DOI: 10.1016/j.autcon.2023.104980.
[5] Mrs. Jain, Sakshi Chaturvedi, Hetvi Patel, and Rajeshwari Jaiswal, “Real-time Face and Object Detection with Age and Gender Prediction for Video Surveillance Applications”, International Journal for Research in Applied Science and Engineering Technology, vol. 11, pp. 2358-2363, 2023. DOI: 10.22214/ijraset.2023.49987.
[6] T. Patel, A. Y. H. Yao, Y. Qiang, W. T. Ooi, and R. Zimmermann, “Multi- Camera Video Scene Graphs for Surveillance Videos Indexing and Retrieval”, in Proceedings of the 2021 IEEE International Conference on Image Processing (ICIP), Anchorage, AK, USA, 2021, pp. 2383- 2387. DOI: 10.1109/ICIP42928.2021.9506713.
[7] Ultralytics, “YOLO: You Only Look One-level”, [Online]. Available: https://www.ultralytics.com/yolo-vision/. [Accessed: Oct. 20, 2024].
[8] Hinepo, “YOLOv8 Finetuning for PPE Detection”, Kaggle, [Online]. Available: https://www.kaggle.com/code/hinepo/ yolov8-finetuning-for-ppe-detection/. [Accessed: Sep. 29, 2023].
[9] N. Yan, H. Cheng, M. Wang, Q. Huang and F. Chen, “DP-Face: Privacy- Preserving Face Recognition Using Siamese Network”, 2021 20th International Symposium on Distributed Computing and Applications for Business Engineering and Science (DCABES), Nanning, China, 2021, pp. 92-95, doi: 10.1109/DCABES52998.2021.00030.
[10] Susanqq, “UTKFace Dataset”, [Online]. Available: https://susanqq. github.io/UTKFace/. [Accessed: Sep. 29, 2023].
