Sistem Identifikasi Brake Lining Berdasarkan Deteksi Lubang Menggunakan Metode You Only Look Once (YOLO)

Pangestu, Pramana Tabah (2026) Sistem Identifikasi Brake Lining Berdasarkan Deteksi Lubang Menggunakan Metode You Only Look Once (YOLO). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2040221024-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (9MB) | Request a copy

Abstract

Brake lining merupakan komponen penting pada sistem pengereman kendaraan yang menuntut kualitas tinggi guna menjamin keselamatan pengguna dan kinerja pengereman yang optimal. Salah satu cacat yang sering muncul dalam proses produksi brake lining yaitu ketidaksesuaian jumlah dan letak lubang pada permukaan produk, sehingga dapat menurunkan kekuatan mekanik, mempercepat keausan, serta mengurangi keandalan fungsi pengereman. Namun, proses inspeksi kualitas di banyak industri otomotif hanya mengandalkan pengamatan manual oleh operator, sehingga hasil inspeksi dipengaruhi oleh keterbatasan kemampuan manusia. Hal ini menyebabkan hasil yang tidak konsisten dan memerlukan waktu yang lama untuk inspeksi kualitas pada lini produksi. Oleh karena itu, diperlukan perancangan untuk implementasi sistem inspeksi kualitas kampas rem menggunakan metode You Only Look Once (YOLO), yang terintegrasi dengan sistem Internet of Things (IoT) dan kontrol otomatis. Sistem ini menggunakan tiga kamera sebagai perangkat akuisisi gambar untuk mendeteksi cacat lubang permukaan pada kampas rem secara real-time, kemudian metode YOLO digunakan sebagai algoritma deteksi objek dengan cepat dan akurat. Hasil deteksi kemudian dikirim ke mikrokontroler ESP32 sebagai perangkat IoT untuk pengambilan keputusan kategorisasi kualitas dan mengontrol aktuator untuk memisahkan produk kampas rem yang diterima (OK) dan yang tidak diterima (NG) secara otomatis. Selain itu, data inspeksi dapat dikirim ke platform IoT untuk tujuan pemantauan, pencatatan buku log kontrol kualitas, dan evaluasi kualitas produksi secara online. Hasil uji menunjukkan sistem ini mampu melakukan deteksi, dengan precision 75,1%, recall 73,5%, mAP50 99,1%, mAP50-95 93,1%, sehingga mampu mengendalikan proses pemisahan produk OK dan NG. =======================================================================================================================================
Brake lining is an important component in a vehicle's braking system that requires high quality to ensure safety and optimal braking performance. A common defect that often occurs in the brake lining production process is the mismatch of the amount and position of holes (voids) on the product, it would reduce mechanical strength, accelerate wear, and decrease the reliability of the braking function. However, the quality inspection process in the many automotive industry only relies on manual observation by operators, so the inspection results are affected by the limitations of human performance. This leads to inconsistent results and requires a long time for quality inspection on the production line. Therefore, a design is needed for the implementation of a brake lining quality inspection system using the You Only Look Once (YOLO) method, integrated with the Internet of Things (IoT) system and automatic control. The system uses three camera for image acquisition device to detect surface hole defects on the brake lining in real-time, then the YOLO method is used as object detection algorithm quickly and accurately. The detection results then sent to the ESP32 microcontroller as an IoT device for making decisions of quality categoriez and controlling actuators to separate acceptable (OK) and unacceptable (NG) brake lining products automatically. Additionally, the inspection data could be sent to the IoT platform for monitoring purposes, recording quality control log book, and online evaluation of production quality. The test results the system is capable of performing detection, and have precision 75,1%, recall 73,5%, mAP50 99,1%, mAP50-95 93,1%, allowing it to control the separation process of OK and NG products reliable.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	brake lining, machine vision, YOLO, IoT, inspeksi kualitas, brake lining, machine vision, YOLO, IoT, quality inspection.
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3070 Automatic control T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK351 Electric measurements. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK6592.A9 Automatic tracking. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7871.674 Detectors. Sensors
Divisions:	Faculty of Vocational > 36304-Automation Electronic Engineering
Depositing User:	Pramana Tabah Pangestu
Date Deposited:	10 Aug 2026 02:32
Last Modified:	10 Aug 2026 02:32
URI:	http://repository.its.ac.id/id/eprint/143013

Actions (login required)

View Item