다중로봇 SLAM을 위한 V2X 통신기반 센서데이터 교환 방법

초록

Multi-robot SLAM requires reliable and timely exchange of information among robots to maintain consistent global mapping, yet existing approaches often face challenges due to limited communication bandwidth, irregular data sharing, and delays in exchanging sensor observations. These communication constraints lead to inconsistencies between independently built maps and reduced mapping reliability in real-world environments. This study proposes a multi-robot SLAM framework in which multiple robots share sensor-level observations, specifically raw LiDAR and IMU data, to enhance global map consistency and examines how real-time data exchange influences SLAM performance in complex environments, allowing each robot to complement its own observations with those of neighboring robots. The contributions of this work include demonstrating the feasibility of sensor-level data sharing for real-world multi-robot SLAM, analyzing its impact on mapping consistency and localization accuracy, and validating a communication-integrated SLAM structure suitable for practical multi-robot operation.