Deteksi Pelanggaran Personal Hygiene Protocol Di Area Produksi Menggunakan Metode Convolutional Neural Network (CNN)

Maulana, Muhammad Irvan Arif (2025) Deteksi Pelanggaran Personal Hygiene Protocol Di Area Produksi Menggunakan Metode Convolutional Neural Network (CNN). Other thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 2040211044-Undergraduate_Thesis.pdf] Text
2040211044-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only

Download (7MB) | Request a copy

Abstract

Sebagai industri pangan yang berfokus pada produksi sosis, PT Charoen Pokphand Indonesia Plant Berbek selalu berusaha untuk menjaga kualitas dari hasil produksi yang dilakukan. Kualitas dari hasil produksi akan mempengaruhi reputasi perusahaan yang erat kaitannya dengan kepercayaan dan daya beli konsumen. Namun, permasalahan berupa pelanggaran terhadap Personal Hygiene Protocol, seperti penggunaan masker yang tidak benar atau tidak memakai jas laboratorium tidak benar. Pelanggaran jas laboratorium tercatat mencapai puncak hingga 820 kasus dalam waktu 8 jam produksi, sementara pelanggaran masker lebih rendah, dengan jumlah lebih dari 5 kasus dalam waktu 8 jam produksi. Hal ini meningkatkan risiko kontaminasi yang dapat berdampak pada kualitas produk dan kepercayaan konsumen. Penelitian ini bertujuan untuk mengatasi permasalahan tersebut dengan mengembangkan sistem deteksi otomatis berbasis Computer Vision menggunakan metode Convolutional Neural Network (CNN) dan arsitektur You Only Look Once (YOLO). Sistem ini dirancang untuk memantau pelanggaran protokol kebersihan secara real-time melalui integrasi kamera CCTV dengan perangkat IoT. Hasil penelitian dari 3 model utama yang diuji adalah YOLOv5, YOLOv8 dan YOLOv11 dengan tipe anotasi polygon menunjukan YOLOv5 polygon menunjukan performa terbaik yang secara konsisten menghasilkan nilai F1-Score lebih tinggi hampir pada semua class, terutama unggul pada wrong coat 95,41%, no hairnet 90,11% dan correct coat 92,78%. Berdasarkan pengujian selama 5 hari produksi dengan waktu pengujian 8 jam setiap harinya, sistem mendeteksi lebih dari 3.150 pelanggaran berupa wrong coat serta 22 kejadian wrong mask. Sistem dapat merekapitulasi setiap jenis pelanggaran dengan waktu terjadinya pelanggaran. Hasil tersebut akan disimpan untuk menyediakan rekapitulasi data pelanggaran sebagai dasar pengambilan tindakan korektif kepada pegawai yang melakukan pelanggaran.
========================================================================================================================
As a food industry focused on sausage production, PT Charoen Pokphand Indonesia Plant Berbek always strives to maintain the quality of its products. Product quality affects the company's reputation, which is closely related to consumer trust and purchasing power. However, there are issues such as violations of the Personal Hygiene Protocol, such as improper use of masks or not wearing lab coats correctly. Violations of laboratory coat usage reached a peak of 820 cases within an 8-hour production period, while mask violations were relatively lower, with over 5 cases within the same 8-hour production period. This increases the risk of contamination, which can impact product quality and consumer trust. This study aims to address these issues by developing an automatic detection system based on Computer Vision using the Convolutional Neural Network (CNN) method and the You Only Look Once (YOLO) architecture. This system is designed to monitor hygiene protocol violations in real-time through the integration of CCTV cameras with IoT devices. The results of testing three main models YOLOv5, YOLOv8, and YOLOv11 with polygon annotation type showed that YOLOv5 polygon performed best, consistently yielding higher F1-Score values across nearly all classes, particularly excelling in wrong coat 95.41%, no hairnet 90.11%, and correct coat 92.78%. Based on testing over 5 days of production with 8 hours of testing each day, the system detected over 3,150 violations involving wrong coats and 22 instances of wrong masks. The system can compile each type of violation along with the time of occurrence. These results will be stored to provide a summary of violation data as the basis for corrective actions against employees who commit violations.

Item Type: Thesis (Other)
Uncontrolled Keywords: Personal Hygiene Protocol, IoT, Convolutional Neural Network (CNN)
Subjects: Q Science > QA Mathematics > QA336 Artificial Intelligence
Q Science > QA Mathematics > QA76.6 Computer programming.
Q Science > QA Mathematics > QA76.625 Internet programming.
Q Science > QA Mathematics > QA76.76.A65 Application software. Enterprise application integration (Computer systems)
Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
T Technology > T Technology (General) > T55 Industrial Safety
T Technology > T Technology (General) > T57.5 Data Processing
T Technology > T Technology (General) > T58.5 Information technology. IT--Auditing
T Technology > T Technology (General) > T58.6 Management information systems
T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing.
Divisions: Faculty of Vocational > 36304-Automation Electronic Engineering
Depositing User: Muhammad Irvan Arif Maulana
Date Deposited: 04 Aug 2025 09:24
Last Modified: 04 Aug 2025 09:25
URI: http://repository.its.ac.id/id/eprint/126732

Actions (login required)

View Item View Item