Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal Jurnal Teknika Widya Teknik
Muhammad Latif
universitas trunodjoyo madura
Chemical Engineering, Chemistry & Bioengineering Civil Engineering, Building, Construction & Architecture Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Design of a Kinematic Control-Based Motion Control System for Path-Following Robots to Support Sustainable Agriculture Muhammad Latif; Lusia Permata S. H
Widya Teknik Vol. 25 No. 2 (2026): November
Publisher : Fakultas Teknik, Universitas Katolik Widya Mandala Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33508/wt.v25i2.8541

Abstract

This study focuses on an autonomous robot that can execute plowing motions as a testing medium in agricultural applications is the main goal of this project. The use of line-following robots as the basis for automated systems has been prompted by the quick development of robotics technology. A line-following robot is appropriate for structured field operations since it is made to follow a predetermined course. In order to meet the need for an automated system that can precisely and reliably follow predetermined paths, a line-following robot prototype is developed in this project. The proposed system utilizes a camera as the primary sensor to detect and determine the robot's position in real time. This visual-based approach enables more flexible and adaptive path detection compared to conventional sensors. To control the robot’s motion, a kinematic control method is implemented. This method is selected because it allows precise regulation of both linear velocity and angular velocity, ensuring that the robot can maintain alignment with the path while executing smooth turning maneuvers. Experimental evaluations were conducted on three different field sizes to assess the performance of the system. The results show that the robot achieved an accuracy of 93.49% on a 2 m² field, 94.49% on a 1.5 m² field, and 95.49% on a 1 m² field. These findings indicate that the proposed kinematic control-based system performs effectively in maintaining path-following accuracy across varying field dimensions. In conclusion, the developed robot demonstrates reliable performance as a line-following system and has potential to be further developed for practical agricultural automation applications, particularly in assisting plowing activities with improved efficiency and consistency.
Evaluasi Komparatif Kinerja YOLOv11 dan ResNet-50 untuk Deteksi Gestur Tangan pada Kendali Differential Drive Mobile Robot Diana Mar’atus Sholihah; Muhammad Latif; Muhammad Fuad; Vivi Tri Widyaningrum; Hairil Budiarto
Jurnal Teknika Vol 18 No 2 (2026): SEPTEMBER
Publisher : Universitas Islam Lamongan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30736/jt.v18i2.1654

Abstract

Pengenalan gestur tangan berbasis visi komputer menawarkan metode interaksi yang lebih intuitif untuk mengendalikan Differential Drive Mobile Robot (DDMR). Penelitian ini bertujuan membandingkan dua pendekatan deep learning yang berbeda, yaitu YOLOv11 sebagai model deteksi objek dan ResNet-50 sebagai model klasifikasi citra. Dataset terdiri atas 500 citra yang terbagi ke dalam lima kelas gestur, yaitu maju, mundur, belok kanan, belok kiri, dan berhenti. Kedua model dilatih menggunakan pendekatan transfer learning dan dievaluasi melalui pengujian internal serta pengujian langsung menggunakan kamera pada sistem kendali robot. Hasil evaluasi internal menunjukkan bahwa YOLOv11 memperoleh nilai mAP50 sebesar 0,995 dan mAP50–95 sebesar 0,784, sedangkan ResNet-50 menghasilkan nilai precision, recall, dan F1-score sebesar 1,000 pada seluruh kelas. Meskipun demikian, hasil pengujian langsung menunjukkan perbedaan yang signifikan. YOLOv11 menghasilkan nilai confidence pada rentang 0,80–0,90 dengan rata-rata 0,844, sedangkan ResNet-50 memperoleh nilai confidence pada rentang 0,32–0,77 dengan rata-rata 0,442. Perbedaan antara evaluasi internal dan pengujian langsung menunjukkan adanya domain gap antara citra dataset dan citra pada kondisi operasional. Berdasarkan skenario pengujian penelitian ini, YOLOv11 menghasilkan nilai confidence yang lebih tinggi dan lebih konsisten sehingga lebih sesuai digunakan untuk pengenalan gestur tangan pada kendali DDMR. Kontribusi penelitian ini terletak pada evaluasi komparatif antara pendekatan deteksi objek dan klasifikasi citra, tidak hanya pada dataset internal, tetapi juga pada penerapan langsung untuk kendali robot berbasis kamera.
Co-Authors Diana Mar’atus Sholihah Hairil Budiarto Lusia Permata S. H Muhammad Fuad Vivi Tri Widyaningrum