Multi-robot real-time collaborative SLAM system based on 5G MEC framework

Aiming at the problem of insufficient computing resources for SLAM calculations in multi-robot systems, a multi-robot collaboration system based on the 5G edge computing platform (MEC) framework is proposed, and a map merge algorithm based on computer vision is designed. First, the main problems of SLAM map construction under multi-robot systems are analyzed, and solutions are given; secondly, a multirobot collaborative SLAM system based on 5G MEC is studied; finally, AKAZE feature detection is used to improve the map merge algorithm to enhance the global map. The algorithm’s performance and the system’s effectiveness are verified by simulation and physical experiments, respectively. The experimental results show that the real-time robot collaboration system based on the 5G MEC framework proposed in this article performs better in map merge. The algorithm has calculation speed and communication delay advantages and will promote research on 5G network applications in real-time robot collaboration.