Rancang bangun sistem robot ARM dengan kendali suara Berbasis Lightweight Transformer Encoder Block (LTEB) dan deteksi warna HSV

Supriadi, Reval (2025) Rancang bangun sistem robot ARM dengan kendali suara Berbasis Lightweight Transformer Encoder Block (LTEB) dan deteksi warna HSV. Sarjana thesis, UIN Sunan Gunung Djati Bandung.

Preview	Text Cover.pdf Download (241kB) | Preview
Preview	Text Abstrak.pdf Download (92kB) | Preview
Preview	Text Keterangan bebas plagiarisme.pdf Download (1MB) | Preview
Preview	Text Daftar isi.pdf Download (100kB) | Preview
Preview	Text Bab 1.pdf Download (96kB) | Preview
	Text Bab 2.pdf Restricted to Registered users only Download (628kB)
	Text Bab 3.pdf Restricted to Registered users only Download (1MB)
	Text Bab 4.pdf Restricted to Registered users only Download (1MB)
	Text Bab 5.pdf Restricted to Registered users only Download (81kB)
	Text Daftar Pustaka.pdf Restricted to Registered users only Download (111kB)
	Text Lampiran.pdf Restricted to Repository staff only Download (423kB)

Abstract

The development of robotic technology demands control systems that are increasingly intuitive, adaptive, and capable of interacting naturally with users. Robot arm control systems that rely on physical buttons or joysticks still face limitations in movement flexibility, response accuracy, and operational efficiency, making them less effective for modern applications requiring dynamic and real-time control. This research proposes the development of a robot arm control system based on voice recognition and color detection as a solution to improve operational efficiency and ease of interaction. The system utilizes the Lightweight Transformer Encoder Block (LTEB) method to classify voice commands into three target colors— red, yellow, and blue—through feature extraction using Log-Mel Spectrogram, while the Hue, Saturation, Value(HSV) color space segmentation method is employed to detect object color and determine its position on a 3×3 grid using centroid calculation. The outputs from both processes are integrated and transmitted as commands to control servo motors via serial communication between Python and Arduino Uno in real-time. Experiments were conducted on 270 test scenarios across three color categories and nine object positions, resulting in a total of five errors consisting of two voice recognition errors for blue commands and three mechanical handling errors during object retrieval. Based on these results, the system achieved a voice recognition accuracy of 99.26%, a mechanical accuracy of 98.89%, and an overall integrated system accuracy of 99.07%. These findings demonstrate that the integration of LTEB and HSV provides stable, effective, and responsive performance for automatic and real-time robot arm control. This research shows potential applications in robotics education, small-scale industrial automation, and voice-controlled assistive tools for individuals with disabilities, and offers opportunities for further development through the addition of Voice Activity Detection (VAD), increased servo torque, and expanded voice datasets to enhance system accuracy.

Actions (login required)

View Item