Mapping Jalan Kawasan Kampus Its Menggunakan Lidar Point Cloud Untuk Data Navigasi Mobil Otonom

Wicaksono, Ivan Manuel (2022) Mapping Jalan Kawasan Kampus Its Menggunakan Lidar Point Cloud Untuk Data Navigasi Mobil Otonom. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 07111840000108-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (2MB)

Abstract

Agar kendaraan otonom dapat berjalan dengan berjalan tanpa campur tangan manusia, kendaraan otonom perlu mengetahui bentuk dari lingkungan yang dilewatinya. Sehingga perlu dibuatkan pemetaan lingkungan agar kendaraan otonom dapat berjalan secara otonom. Pembuataan pemetaan sendiri diperlukan alat dan bahan yaitu LiDAR 3D RS-16 dengan akurasi 2 cm, Pixhawk 2.1, GPS Module ZED-F9P, Telemetry RFD 900x dengan jangkauan lebih dari 15 km. Agar mendapatkan lokasi GPS yang akurat, diperlukan penambahan RTK/GPS reference sehingga terjadi komunikasi serial antara GPS Base Module dengan GPS Rover Module dengan akurasi dapat mencapai 10 mm. Dalam penelitian ini pembuatan peta diperlukan penerapan algoritma yaitu Karto SLAM untuk pembuatan peta dan metode RANSAC Plane Segmentation untuk mencari bentuk jalan. Pengambilan dataset dilakukan di kawasan ITS yaitu bundaran robotika hingga Research Center. Evaluasi dilakukan dengan membandingkan hasil pengukuran pada Rviz dengan jarak pada Google Maps. Bentuk peta sendiri terdapat beberapa bagian yang hilang saat putar balik dikarenakan tidak ada halangan yang terdeteksi oleh I-CAR Dari hasil evaluasi penelitian didapatkan error rata-rata jarak dari titik ke titik pada peta sebesar 8.4926%, error rata-rata lebar jalan pada peta sebesar 2.0008%, sedangkan akurasi perbandingan area menggunakan Intersection Over Union (IOU) sebesar 71.48%.
==============================================================================================================================
In order for autonomous vehicles to run without human intervention, autonomous vehicles need to know the shape of the environment they pass through. So it is necessary to make environmental mapping so that autonomous vehicles can run autonomously. The mapping itself requires tools and materials, namely LiDAR 3D RS-16 with 2 cm accuration, Pixhawk 2.1, GPS Module ZED-F9P, Telemetry RFD 900x with range over 15 km. In order to get an accurate GPS location, it is necessary to add an RTK/GPS reference so that serial communication occurs between the GPS Base Module and the GPS Rover Module that is capabale of 10 mm accuration. In this research, it is necessary to apply an algorithm, namely Karto SLAM for map making and the RANSAC Plane Segmentation method to find the shape of the road. Dataset collection was carried out in the ITS area, namely the robotics roundabout to the Research Center. The evaluation is done by comparing the measurement results on Rviz with the distance on Google Maps. In the shape of the map itself, there are several parts that are lost when turning around because there are no obstacles detected by I-CAR. The results of the research evaluation, the average error of the distance from point to point on the map is 8.4926%, the average error of road width on the map is 2.0008%, while the accuracy of area comparison using Intersection Over Union (IOU) is 0.7148%.

Item Type:	Thesis (Other)
Additional Information:	RSE 629.893 2 Wic m-1 2022
Uncontrolled Keywords:	Autonomous Car. LiDAR. GPS. SLAM. Point cloud.
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User:	Mr. Marsudiyana -
Date Deposited:	12 Jun 2026 06:50
Last Modified:	12 Jun 2026 06:50
URI:	http://repository.its.ac.id/id/eprint/133774

Actions (login required)

View Item