Abstract Navigating mobile robots through dynamic and cluttered environments presents a significant challenge, particularly for applications in service and logistics, which demand both precise path tracking and adaptive autonomy. Conventional navigation methods, which often rely on physical markers or static environmental maps, typically exhibit limited robustness and adaptability. This paper introduces a comprehensive design and implementation of a dual-map navigation system within the Robot Operating System (ROS) framework, which incorporates a virtual obstacle layer to enhance safety and control. The system leverages Simultaneous Localization and Mapping (SLAM) for real-time map generation and strategically decouples localization from path planning by employing distinct maps for each task. An original, unmodified map ensures accurate robot pose estimation, while a secondary, modified map augmented with user-defined or algorithmically-generated virtual obstacles constrains the robot's navigable space. Extensive experiments conducted with the xbot-u mobile platform in complex laboratory settings demonstrate significant improvements in navigation accuracy, robustness, and safety. The proposed method effectively prevents collisions with obstacles that are imperceptible to 2D sensors, such as tabletops. This highly adaptable, markerless, and scalable solution offers a practical pathway for deploying robots in real-world, dynamic environments. Future work will focus on integrating semantic information and advanced dynamic obstacle management to further enhance system intelligence.
Zhou et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: