Crowd OS Powered Active 3D Mapping Leveraging Heterogeneous Crowd Robots

This paper presents a method for active 3D mapping leveraging heterogeneous crowd robots. By incorporating a UAV with its aerial perspective and powerful perception capabilities, the blind spots present in ground robot mapping are successfully eliminated, resulting in more comprehensive and accurate mapping outcomes. Furthermore, our heterogeneous crowd robots exhibits high scalability, allowing for the addition of a specified number of robots based on the mapping requirements of different indoor environments, while also supporting dynamic addition and removal. This scalability enables the system to cater to diverse applications and establishes a foundation for a more robust, comprehensive, and sustainable collective robot system. Through this research, we fully demonstrate the advantages of integrating heterogeneous robotic systems and how different types of robots with complementary capabilities collaborate to improve indoor 3D mapping processes. In particular, the introduced CrowdOS plays a significant role in task assignment and coordination, offering powerful functionalities for intelligent task allocation and coordination among different robots.