Sistem Deteksi Jalan Berlubang Berbasis CNN untuk Mendukung Peningkatan Keselamatan Berkendara dan Kualitas Jalan

Arif, Muhammad Hasyim Al (2026) Sistem Deteksi Jalan Berlubang Berbasis CNN untuk Mendukung Peningkatan Keselamatan Berkendara dan Kualitas Jalan. Undergraduate thesis, Politeknik Negeri Jember.

This is the latest version of this item.

	Text (Abstract) Abstract.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (42kB)
	Text (Bab 1 Pendahuluan) Bab 1 Pendahuluan.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (126kB)
	Text (Daftar Pustaka) Daftar Pustaka.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (376kB)
	Text (Laporan Lengkap) Laporan Lengkap.pdf Restricted to Registered users only Download (5MB) \| Request a copy

Abstract

Road damage, particularly potholes, is a major obstacle to mobility and increases the risk of traffic accidents, thereby creating a need for intelligent and rapid monitoring technology. This study aims to design and implement an automatic road damage detection system using a Convolutional Neural Network (CNN) algorithm with the YOLOv8 model integrated with a Geographic Information System (GIS). Using the Waterfall development method, the system was implemented on a Raspberry Pi 5 device, a Neo M8N GPS module, as well as the SiALAN website platform and SiAlan Apps mobile application. The model is capable of identifying five categories of road damage, namely potholes, alligator cracks, edge cracks, patching, and weathering/raveling. The results indicate a high level of accuracy, ranging from 88% to 93%, with the alligator cracking category achieving the best performance, recording an F1-Score of 82.1%. Further testing demonstrated that vehicle speed affects detection effectiveness, with optimal results achieved at a speed of 10 km/h, yielding an average detection success rate of 84%. Meanwhile, at a speed of 30 km/h, performance decreased significantly to 40% due to the motion blur effect. Furthermore, the User Acceptance Test (UAT) results showed a system feasibility score of 84%, which falls into the Very Good or Highly Acceptable category. Overall, the developed system provides easy access to accurate information on the location and type of road damage, thereby supporting monitoring, decision-making, and road infrastructure maintenance planning. As a result, the system serves as an effective tool for improving driving safety and road quality

Item Type:

Thesis (Undergraduate)

Contributors: