SAFEGUARD: AI-Powered Industrial Safety Monitoring System

In this project, an AI-Based Industrial Worker Health and Safety Monitoring System has been proposed using Computer Vision and Artificial Intelligence technologies. The system was developed to improve workplace safety by continuously monitoring industrial workers and detecting safety violations in real time.   The proposed framework uses a camera-based monitoring system to detect workers and identify whether mandatory Personal Protective Equipment (PPE) such as helmets and safety vests are being used properly. The system also includes a Hazard Zone Monitoring feature that detects unauthorized worker entry into dangerous or restricted industrial areas. Whenever safety violations are detected, the system activates a siren alert and automatically sends email notifications to the manager or supervisor for immediate action.   The implementation demonstrated that the proposed system can effectively reduce dependency on manual supervision and provide an efficient automated solution for industrial safety management. The preprocessing and object detection modules helped improve monitoring accuracy and real-time detection performance under different environmental conditions.   Compared to traditional manual monitoring systems, the proposed framework provides a smarter, faster and more cost-effective approach for workplace safety monitoring. The developed prototype also shows strong potential for future implementation in smart industries and Industry 4.0 environments.   Overall, the proposed Industrial Worker Health and Safety Monitoring System contributes toward improving worker protection, reducing accident risks and enhancing industrial automation using Artificial Intelligence, Computer Vision and intelligent alert mechanisms

1.      Redmon, J., Divvala, S., Girshick, R., and Farhadi, A., “You Only Look Once: Unified, Real-Time Object Detection,” IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 779–788, 2016. This paper introduced the YOLO algorithm for fast and real-time object detection. 2.      Bochkovskiy, A., Wang, C. Y., and Liao, H. Y. M., “YOLOv4: Optimal Speed and Accuracy of Object Detection,” arXiv preprint arXiv:2004.10934, 2020. This research improved object detection accuracy and speed using the YOLOv4 model. 3.      Jocher, G., et al., “YOLO by Ultralytics,” GitHub Repository, 2023.This repository provides practical implementation of YOLO models for real-time object detection applications. 4.      Viola, P., and Jones, M., “Rapid Object Detection using a Boosted Cascade of Simple Features,” IEEE Conference on Computer Vision and Pattern Recognition (CVPR), vol. 1, pp. 511–518, 2001. This paper introduced a fast object detection method using Haar Cascade classifiers. 5.      Bradski, G., “The OpenCV Library,” Dr. Dobb’s Journal of Software Tools, 2000.This work introduced OpenCV, an important library for image processing and computer vision tasks. 6.      Ren, S., He, K., Girshick, R., and Sun, J., “FasterR-CNN: Towards Real-Time Object Detection with Region Proposal Networks,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 39, no. 6, pp. 1137–1149, 2017.This research proposed Faster R-CNN for accurate and efficient object detection. 7.      He, K., Gkioxari, G., Dollár, P., and Girshick, R., “Mask R-CNN,” IEEE International Conference on Computer Vision (ICCV), pp. 2961–2969, 2017. This paper presented Mask R-CNN for object detection and image segmentation. 8.      Sultana, F., Sufian, A., and Dutta, P., “Advancements in Image Classification using Convolutional Neural Network,” International Conference on Computer Engineering and Technology, pp. 122–129, 2018. This study discussed the effectiveness of CNN models in image classification applications. 9.      Girshick, R., “Fast R-CNN,” IEEE International Conference on Computer Vision (ICCV), pp. 1440–1448, 2015.This paper introduced Fast R-CNN for improving object detection speed and performance. 10.   Wang, Z., Qin, Z., and Zhang, Y., “Real-Time Helmet Detection Based on Deep Learning for Worker Safety,” IEEE Access, vol. 8, pp. 196650–196660, 2020.This research focused on deep learning-based helmet detection for industrial worker safety monitoring..