Although voice communication plays a crucial role in maritime operations due to its safety and efficiency, Maritime Autonomous Surface Ships (MASS) operating unmanned have not been fully integrated with existing voice communication systems. Therefore, this study proposes an integrated voice communication system for MASS. The proposed system combines state-of-the-art Speech-to-Text (STT) technology, Text-to-Speech (TTS) technology, and Large Language Models (LLM) to enable MASS to autonomously interact via VHF voice communication. In particular, the LLM generates appropriate responses that comply with standardized maritime communication protocols by learning Standard Marine Communication Phrases (SMCP) using Retrieval-Augmented Generation (RAG). The developed system enhances the autonomy of MASS and enables seamless interaction with existing manned ships and control centers. Experimental results demonstrated high speech recognition accuracy and real-time processing performance, confirming its practicality in maritime communication environments.
AI-Driven Conversational Voice Communication for Maritime Autonomous Surface Ships
2025-02-18
260517 byte
Conference paper
Electronic Resource
English
Data/Knowledge-Driven Behaviour Analysis for Maritime Autonomous Surface Ships
TIBKAT | 2023
|Data/Knowledge-Driven Behaviour Analysis for Maritime Autonomous Surface Ships
GWLB - Gottfried Wilhelm Leibniz Bibliothek | 2023
|