Wireless Multi-Robot Collaboration: Communications, Perception, Control, and Planning

The Internet of Robots integrates robotic devices with wireless networks and cloud computing, which is widely applied in manufacturing, autonomous driving, agriculture and so on. However, The communication capabilities of robots are constrained in complex physical environments, impeding the advancement of large-scale multi-robotic collaboration. An experimental platform that seamlessly integrates communication, perception, control, and planning is necessary for robotic collaborative algorithm implementation and evaluation. In this paper, we propose an effective multi-robot wireless network collaboration architecture and a novel semi-physical simulation experimental platform incorporating the physical environment, electromagnetic wave propagation, wireless communication transmission, and networking topology. We employ the three-dimensional (3D) ray tracing model to simulate electromagnetic wave propagation and calculate network transmission characteristics based on coverage analysis. To perform a semi-physical simulation, our collaboration system utilizes a semantic segmentation model to enhance robotic perception, in which an adaptive wireless Data Distribution Service (DDS) protocol is used for communication among robots and servers that execute global control and planning. The results demonstrate that the system can achieve centimeter-level precision in perception, reduce task time by up to 73.6% when increasing from one to four robots, and decrease computational utilization by 5.0% for robots.