Sistem Pendeteksi Kerusakan Jahitan Kantong Pupuk untuk Menurunkan Kerugian Pupuk Tumpah Menggunakan Metode Convolutional Neural Network Pada Line Pengantongan Urea

Nasrulloh, Fikri Ahmad Dwi (2025) Sistem Pendeteksi Kerusakan Jahitan Kantong Pupuk untuk Menurunkan Kerugian Pupuk Tumpah Menggunakan Metode Convolutional Neural Network Pada Line Pengantongan Urea. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2040211025-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (10MB) | Request a copy

Abstract

PT Petrokimia Gresik (PKG), anak perusahaan PT Pupuk Indonesia, bertanggung jawab atas produksi dan distribusi pupuk subsidi di Indonesia, dengan salah satu produk utamanya adalah pupuk urea. Proses pengantongan urea yang dilakukan di Gudang Urea mengalami peningkatan permintaan yang signifikan, namun hal ini juga membawa tantangan terkait kerugian produk akibat kerusakan jahitan kantong. Kerusakan jahitan, seperti jahitan yang tidak sempurna, dapat menyebabkan pupuk berceceran, yang berpotensi menimbulkan kerugian material. Meskipun pengawasan kualitas jahitan sudah dilakukan melalui cctv proses ini sering kali lalai terhadap kerusakan jahitan dalam volume produksi yang tinggi. Penelitian ini bertujuan mengembangkan sistem deteksi otomatis berbasis Convolutional Neural Network (CNN) dengan arsitektur You Look Only Once V8 untuk mendeteksi kerusakan jahitan kantong pupuk secara real-time pada 4 line conveyor pengantongan. Sistem ini akan mendeteksi cacat jahitan, pada proses pengisian produk dengan akurasi sebesar 91,2%. Menggunakan dataset citra jahitan kantong yang diambil langsung dari proses operasional di gudang pengantongan urea, model ini dilatih dengan teknik augmentasi untuk meningkatkan akurasi deteksi. Implementasi sistem ini dapat mengurangi kerugian produk akibat pupuk yang berceceran atau tumpah sebesar 77% dan memastikan kualitas produk yang lebih baik dengan meminimalkan distribusi kemasan yang tidak sempurna sebesar 80%.
========================================================================================================================
PT Petrokimia Gresik (PKG), a subsidiary of PT Pupuk Indonesia, is responsible for the production and distribution of subsidized fertilizers in Indonesia, with one of its main products being urea fertilizer. The urea bagging process in the Urea Warehouse has experienced a significant increase in demand, but this has also brought challenges related to product losses due to bag stitching damage. Stitching defects, such as imperfect seams, can cause fertilizer to spill, potentially leading to material losses. Although quality control of stitching has been carried out through CCTV, this process often overlooks stitching defects in the high-volume production. This study aims to develop an automatic detection system based on Convolutional Neural Network (CNN) with a You Only Look Once V8 architecture to detect stitching defects in fertilizer bags in real-time on four bagging conveyor lines. This system will detect stitching flaws during the product filling process with an accuracy of 91.2%. Using a bag stitching image dataset directly taken from the operational process in the urea bagging warehouse, the model is trained with augmentation techniques to improve detection accuracy. The implementation of this system can reduce product loss due to spilled or spilled fertilizer by 77% and ensure better product quality by minimizing the distribution of imperfect packaging by 80%.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Convolutional Neural Network, Pendeteksi cacat jahitan, You Look Only Once
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing. T Technology > TK Electrical engineering. Electronics Nuclear engineering T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3070 Automatic control
Divisions:	Faculty of Vocational > 36304-Automation Electronic Engineering
Depositing User:	Fikri Ahmad Dwi Nasrulloh
Date Deposited:	04 Aug 2025 09:12
Last Modified:	04 Aug 2025 09:12
URI:	http://repository.its.ac.id/id/eprint/127204

Actions (login required)

View Item