국내기사
엣지 AI 기반의 청음드론(A-LiD)을 활용한 전술적 적 침투 조기 경보 체계 연구 = A study on early warning of enemy infiltration activities using an Edge-AI-based acoustic listening drone
This study proposes an Acoustic-Landed Intelligent Drone (A-LiD) system that mitigates ego-noise in drone-based acoustic reconnaissance by adopting a landing and motor shut-off operational concept, enabling the drone to function as a stationary sensor with significantly improved SNR. Based on this concept, effective acoustic detection ranges are defined according to target type: 60–80 m for dismounted personnel, 200–300 m for small vehicles, and 500 m–1 km for large maneuvering equipment, with the primary operational objective focused on identifying stealthy infiltrating personnel within a 100 m radius. The proposed system processes acoustic data on-site using Edge AI and swarm-based triangulation, offering a practical enhancement to survivability and surveillance efficiency within the Manned–Unmanned Teaming (MUM-T) framework.