This project focuses on developing a multi-terminal collaborative situation understanding system utilizing semantic communication technology and its application to firefighting activities. We aim to establish technologies that support situation awareness and decision-making in emergency situations while efficiently utilizing limited radio resources.
Semantic communication is a communication method that optimizes transmission based on the "meaning" of information. While conventional communication aims to transmit data accurately, semantic communication focuses on ensuring that the intended meaning and purpose are correctly understood.
Each terminal uses AI to understand the meaning of information and transmits only necessary information according to its importance. For example, in transmitting disaster footage, conventional methods send the entire video, whereas semantic communication sends only situational summaries such as "building collapsed, injured persons present." The receiver understands the situation from this summary using a common knowledge base.
This approach enables efficient aggregation of information from multiple terminals even in bandwidth-limited communication environments, achieving significant reductions in communication volume, lower latency, and energy savings. Beyond disaster response, it is expected to be applied to 6G communications, AI-to-AI communication, autonomous driving, and other areas.