Sistem Deteksi Kondisi Baut Engine Part Pada Jalur Perakitan Menggunakan Metode Convolutional Neural Network untuk Menjaga Kualitas Produk

Salsabila, Poppy Putri (2024) Sistem Deteksi Kondisi Baut Engine Part Pada Jalur Perakitan Menggunakan Metode Convolutional Neural Network untuk Menjaga Kualitas Produk. Diploma thesis, Institut Teknologi Sepuluh Nopember.

Text 2040201101_Undergraduate_Thesis for D4 Program.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (5MB) | Request a copy

Abstract

Proses inspeksi merupakan salah satu prosedur yang penting dalam sebuah rangkaian proses produksi di perusahaan otomotif. Hal ini bertujuan agar kualitas produk terjaga sampai ke konsumen, karena apabila produk sampai ke konsumen tidak sesuai dengan standar perusahaan maka akan mengakibatkan kerugian bagi konsumen dan perusahaan. Kerugian yang didapat konsumen seperti kebocoran pada part yang disebabkan tidak terpasangnya baut pada salah satu bagian engine part atau tidak berfungsinya part apabila terdapat benda asing yang menempel. Bagi perusahaan potensi kerugian didapat apabila terdapat keluhan dari konsumen dan turunnya kepercayaan masyarakat pada perusahaan akibat kesalahan proses pengecekan kualitas. Menurut data perusahaan, setiap bulannya terdapat temuan kesalahan inspeksi pada engine, rata-rata Defect Per Unit (DPU) dari bulan Mei sampai bulan November 2023 sebesar 4.08%. Kesalahan disebabkan oleh beberapa faktor dalam melakukan inspeksi berupa kesalahan ketelitian karena banyaknya baut, kelalaian dari operator berupa ketidak sengajaan terlewatnya pemasangan baut atau tertempelnya benda asing, dan lupa lokasi penempatan baut pada engine. Untuk itu, diperlukan adanya sistem yang dapat membantu dalam proses inspeksi baut pada engine part. Sistem yang dibuat mengaplikasikan metode Convolutional Neural Network. Pemilihan metode ini karena kemampuannya dalam melakukan pengenalan dan klasifikasi objek yang terdiri dari beberapa fitur seperti convolutional layer, pooling layer dan fully connected layer. Sistem yang dibuat dapat mendeteksi terpasanganya baut pada engine part dan mendeteksi apabila ditemukan adanya temuan benda asing, dengan tingkat akurasi yang dihasilkan dari sistem yang dibuat sebesar 96%. Dengan demikian, adanya implementasi sistem ini mengatasi terjadinya hambatan pada proses produksi selanjutnya, menjadikan proses produksi terutama di lini produksi berjalan sesuai dengan jadwal, dan terjaganya kualitas engine part
=================================================================================================================================
The inspection process is one of the important procedures in a series of production processes in automotive companies. It aims to maintain product quality to consumers, because if the product reaches consumers not in accordance with company standards, it will result in losses for consumers and companies. Losses obtained by consumers such as leaks in parts caused by not installing bolts on one part of the engine part or not functioning parts if there are foreign objects attached. For the company, potential losses are obtained if there are complaints from consumers and a decrease in public trust in the company due to errors in the quality checking process. According to company data, every month there are findings of inspection errors on the engine, the average Defect Per Unit (DPU) from May to November 2023 is 4.08%. Errors are caused by several factors in conducting inspections in the form of errors in accuracy due to the large number of bolts, negligence from operators in the form of accidentally missing the installation of bolts or sticking foreign objects, and forgetting the location of bolt placement on the engine. For this reason, a system is needed that can assist in the bolt inspection process on engine parts. The system created applies the Convolutional Neural Network method. The selection of this method is due to its ability to recognize and classify objects consisting of several features such as convolutional layer, pooling layer and fully connected layer. The system created can detect the pairing of bolts on engine parts and detect if there are findings of foreign objects, with an accuracy level resulting from the system created being 96%. Thus, the implementation of this system overcomes the occurrence of obstacles in the next production process, making the production process, especially in the production line, run according to schedule, and maintaining the quality of engine parts.

Item Type:	Thesis (Diploma)
Uncontrolled Keywords:	Convolutional Neural Network, Inspeksi Baut, Kualitas Produk, Bolt Inspection, Product Quality
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T59.7 Human-machine systems. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3070 Automatic control T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.546 Computer algorithms T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7888.3 Digital computers
Divisions:	Faculty of Vocational > 36304-Automation Electronic Engineering
Depositing User:	Poppy Putri Salsabila
Date Deposited:	19 Aug 2024 01:06
Last Modified:	19 Aug 2024 01:06
URI:	http://repository.its.ac.id/id/eprint/115430

Actions (login required)

View Item