Research on Improving Self-Localization of Cloud-Assisted Autonomous Mobile Robots

Fusion Method and Evaluation of AMCL and Fast-LIO

Abstract

Against the backdrop of high-speed networks such as post-5G, the introduction of cloud robotics to outdoor working mobile robots is gaining attention. In this study, we propose a system that fuses the robot's high-speed but less robust 2D self-localization with the cloud's high-precision, robust but latency-prone 3D self-localization. This approach aims to achieve a robust self-localization that is both accurate and resilient to communication failures, while balancing cost and computational resources.