Implementasi Sistem Deteksi dan Perhitungan Karung Pupuk Menggunakan YOLOv12n pada Proses Pemuatan di Gudang Multi Guna

Ghanim, Ammara Fazila (2026) Implementasi Sistem Deteksi dan Perhitungan Karung Pupuk Menggunakan YOLOv12n pada Proses Pemuatan di Gudang Multi Guna. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2040221098-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (5MB) | Request a copy

Abstract

Proses distribusi pupuk di Gudang Multi Guna PT Petrokimia Gresik saat ini masih menggunakan penghitungan jumlah karung secara manual oleh operator saat pemuatan ke dalam truk. Metode konvensional ini memiliki keterbatasan berupa risiko kelelahan operator, subjektivitas, dan inkonsistensi pengamatan yang dapat menyebabkan ketidakakuratan data. Berdasarkan data operasional periode April hingga November 2025, tercatat akumulasi selisih muatan sebesar 33,19 ton atau setara dengan 664 karung, dengan tingkat penyimpangan rata-rata sebesar 1,74%. Untuk mengatasi permasalahan tersebut, pada penelitian ini menerapkan Smart Docking System, yaitu sistem penghitungan otomatis menggunakan model YOLOv12n untuk mendeteksi objek karung pupuk dan palet secara real-time dari sudut pandang atas (high-angle view). Hasil deteksi selanjutnya diproses menggunakan algoritma PolygonZone untuk menentukan dan menghitung objek yang memasuki area penghitungan. Hasil penghitungan ditampilkan melalui Graphical User Interface (GUI) dan disimpan secara otomatis ke dalam file lokal. Sistem juga diintegrasikan dengan mikrokontroler ESP32 untuk mengaktifkan alarm suara secara otomatis berdasarkan kondisi jumlah muatan. Berdasarkan hasil pengujian pada lima sesi pemuatan dengan jumlah aktual sebanyak 160 karung pada setiap sesi, sistem memperoleh rata-rata akurasi penghitungan sebesar 92,50%.
==================================================================================================================================
The fertilizer distribution process at PT Petrokimia Gresik’s Multipurpose Warehouse currently still relies on operators manually counting the number of bags during loading onto trucks. This conventional method has limitations, including the risk of operator fatigue, subjectivity, and inconsistencies in observations, which can lead to data inaccuracies. Based on operational data from April through November 2025, a cumulative loading discrepancy of 33.19 metric tons equivalent to 664 bags was recorded, with an average deviation rate of 1.74%. To address these issues, this study implements a Smart Docking System an automated counting system using the YOLOv12n model to detect fertilizer bags and pallets in real time from a high-angle view. The detection results are then processed using the PolygonZone algorithm to identify and count objects entering the counting area. The counting results are displayed via a Graphical User Interface (GUI) and automatically saved to a local file. The system is also integrated with an ESP32 microcontroller to automatically trigger an audible alarm based on the load quantity. Based on test results from five loading sessions each with an actual count of 160 bags the system achieved an average counting accuracy of 92.50%.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Computer Vision, PolygonZone, YOLOv12n,Computer Vision, PolygonZone, YOLOv12n
Subjects:	Q Science > QA Mathematics > QA336 Artificial Intelligence Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) 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 > TK6592.A9 Automatic tracking. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.P3 Pattern recognition systems
Divisions:	Faculty of Vocational > 36304-Automation Electronic Engineering
Depositing User:	Ammara Fazila Ghanim
Date Deposited:	06 Aug 2026 04:25
Last Modified:	06 Aug 2026 04:25
URI:	http://repository.its.ac.id/id/eprint/144143

Actions (login required)

View Item