DESIGN AND IMPLEMENTATION OF NATURAL LANGUAGE PROCESSING AND WEB SERVER INTERFACE FOR THE VOICE INTERACTION BASED COMMUNICATION MODULE ON LOGISTIC ROBOTS

• Main Author: Sangmahata Purba, Harry
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/66407
• Institution: Institut Teknologi Bandung

ABSTRACT DESIGN AND IMPLEMENTATION OF NATURAL LANGUAGE PROCESSING AND WEB SERVER INTERFACE FOR THE VOICE INTERACTION BASED COMMUNICATION MODULE ON LOGISTIC ROBOTS By Harry Sangmahata Purba NIM: 13218002 (ELECTRICAL ENGINEERING PROGRAM) Robot technology continues to develop to help humans complete their work. This type of robot is usually called a service robot. One of the service robots being developed at the Control System and Computer Laboratory of ITB is a delivery robot or logistics robot. This robot does not yet have a communication module, so it is necessary to design and implement a communication module that can increase user interactivity with the robot. Interactivity between the user and the robot must include aspects of friendship and social presence to increase interactivity. One method of communication that uses these aspects is to use speech. For this reason, a voice interaction-based communication module has been designed, VoTion: Voice Interaction in Motion. VoTion uses Natural Language Processing (NLP) to process human voices into text that the system can understand. The text will then be processed into a voice response output, the status of the robot that will be displayed on the interface, and the robot’s movement. In designing the NLP of this system, it takes a part of the system that can perform Speech-to-Text and Text-to-Speech. The framework that will be used to perform Speech-to-Text is Coqui-STT. The framework requires a language model that is obtained from training. The Text-to-Speech framework used is Google Text-to-Speech (GTTS). GTTS can produce sounds in Indonesian that are close to human speech. In the Interface Design of the system, HTTP communication protocol and MySQL database will be used. Keywords: service robot, voice interaction, interface, Speech-to-Text, Text-to-Speech