Implementasi Yolov8 Terhadap Kondisi Wajah Tertutup Sebagian (Occulusion) Untuk Deteksi Dini K3 Di Hanggar Perakitan PT. Industri Kereta Api

Harly, Daniswara Andhika Putra (2026) Implementasi Yolov8 Terhadap Kondisi Wajah Tertutup Sebagian (Occulusion) Untuk Deteksi Dini K3 Di Hanggar Perakitan PT. Industri Kereta Api. S1 thesis, Universitas PGRI Madiun.

	Text Halaman Depan (Cover).pdf Download (975kB)
	Text Abstrak.pdf Download (182kB)
	Text Bab I.pdf Download (235kB)
	Text Bab II.pdf Download (429kB)
	Text Bab III.pdf Restricted to Repository staff only Download (286kB)
	Text Bab IV.pdf Restricted to Repository staff only Download (1MB)
	Text Bab V.pdf Download (222kB)
	Text Daftar Pustaka.pdf Download (199kB)
	Text Lampiran.pdf Restricted to Repository staff only Download (1MB)

Abstract

The assembly hangar area of PT Industri Kereta Api (PT INKA) is a high-risk work area that requires strict implementation of Occupational Safety and Health (K3) protocols, including the mandatory use of Personal Protective Equipment (PPE) such as helmets, masks, and safety glasses. However, the use of these attributes creates a condition of partially covered faces (occlusion), which often makes it difficult to monitor K3 compliance manually due to limited personnel and the size of the hangar area. This study aims to implement a computer vision-based K3 early detection system using the robust YOLOv8 Deep Learning algorithm for occlusion conditions at PT INKA. This system is developed in the form of a web-based monitoring dashboard application integrated with hardware in the form of a camera (webcam) as a real-time digital image capture. Model testing was carried out using a dataset of worker images with various variations of facial PPE attribute occlusion. The results show that the YOLOv8 model is able to detect and classify the compliance status of PPE use in partially covered face conditions in real-time with an accuracy level of mean Average Precision (mAP). The developed website-based system successfully visualizes detection results, records violation history, and automatically sends early warning alerts to safety officers. Therefore, this system can be an effective preventative solution to optimize occupational safety oversight while supporting the achievement of the Zero Accident target in the railway manufacturing industry.

Item Type:	Thesis/Skripsi/Tugas Akhir (S1)
Kata Kunci:	YOLOv8, Alat Pelindung Diri (APD), Oklusi (Occlusion), Deteksi Dini K3, Dashboard Monitoring, PT INKA YOLOv8, Personal Protective Equipment (PPE), Occlusion, OHS Early Detection, Monitoring Dashboard, PT INKA
Subjects:	T Technology > T Technology (General) T Technology > TA Engineering (General). Civil engineering (General) T Technology > TH Building construction T Technology > TS Manufactures
Divisions:	Fakultas Teknik > Teknik Informatika
Depositing User:	PUTRA ANDHIKA DANISWARA
Date Deposited:	30 Jul 2026 05:35
Last Modified:	30 Jul 2026 05:35
URI:	http://eprint.unipma.ac.id/id/eprint/8254

Actions (login required)

View Item