Application of Automatic Detection of Personal Protective Equipment (PPE) for Loading and Unloading Workers at PT. Pelindo TPK Makassar

The purpose of this study is to apply the use of an automatic PPE (Personal Protective Equipment) detection system prototype in supporting the Health, Safety, Security, and Environment (HSSE) efforts of PT. Pelindo Terminal Peti Kemas Makassar also known as TPK Makassar to improve the effectiveness of supervision and reduce the risk of workplace accidents. Workplace accidents can be caused by various factors, ranging from worker negligence in not using PPE, lack of discipline, and carelessness. Safety and health are top priorities for workers and companies. Field testing results show that the system can consistently distinguish between workers who comply with and those who violate PPE usage rules, and send automatic alerts when violations occur. The YOLOv8-based detection model shows fairly good performance with a precision value of 0.7–0.9, recall of 0.5–0.6, and mAP50 reaching 0.7–0.8, which indicates that the system is capable of recognizing main objects with high accuracy.