<![CDATA[Abstract: Military reconnaissance operations in urban warfare environments demand high safety standards for infantry personnel. However, conventional Unmanned Ground Vehicles (UGVs) are often hindered by teleoperation latency and poor standard GPS accuracy in Non-Line-of-Sight areas. This study aims to develop a semi-autonomous hybrid navigation system integrating a Multi-Constellation GNSS module (u-blox Neo-M9N) with Raspberry Pi 5-based Edge Computing technology. The research method employs an experimental engineering design approach to validate navigation performance through static and dynamic testing. The results demonstrate that Multi-GNSS integration significantly improves positioning precision, reducing the Root Mean Square Error (RMSE) by 73.5% from 3.10 meters to 0.82 meters compared to standard GPS modules. Furthermore, the utilization of Raspberry Pi 5 successfully cut navigation algorithm processing latency from 207 ms (on ESP32) to 7.5 ms, effectively minimizing blind distance during high-speed maneuvers. This system proves capable of enhancing reconnaissance operation reliability with minimal path deviation, offering a precise and responsive tactical solution to support situational awareness in conflict zones. Keywords: Edge Computing, Military Reconnaissance Robot, Multi-GNSS, Hybrid Navigation, Raspberry Pi 5.]]>
Copyrights © 2026
