AI Radio System for Unmanned Surface Vessels
The Chair of Applied Computer Science at TU Dortmund University is developing an AI-powered radio system for unmanned surface vessels (USVs). This system aims to improve communication among autonomous systems, particularly in high-traffic areas such as ports and inland waterways.
At the heart of the system is a fully integrated large language model that operates without an internet connection. It enables USVs to receive radio messages and automatically respond using an AI voice. The system can be integrated into existing hardware and communicates directly with the unmanned vehicle’s control system. This allows USVs to carry out instructions from other traffic participants, such as performing evasive maneuvers. Development of the system is currently ongoing.
The system addresses a key challenge: the safe and efficient integration of autonomous vehicles into complex traffic environments. The ability to communicate via radio automatically and to carry out instructions significantly increases the operational safety of USVs. This is particularly important for applications in logistics, surveillance, and research on waterways, where interaction with manned vessels and infrastructure is essential. The system’s independence from an internet connection also makes it robust and versatile, even in remote or infrastructure-poor areas.
