An Adaptive ROS2 Node Deployment Framework in Mobile Edge-Robot Systems

Abstract

Mobile edge computing is an emerging computing paradigm that enhances the computational capabilities of mobile devices by offloading intensive tasks to edge servers. In robotic systems, MEC can significantly reduce response times and improve user experience. However, as robots move through their environments, factors such as the distance to edge servers and physical obstructions fluctuate, leading to variations in communication bandwidth and, consequently, communication delays. To address these challenges, this paper proposes an adaptive computing node offloading framework (ARDF) designed to optimize the dynamic deployment of Robot Operating System 2 (ROS2) nodes in robot-edge environments. The framework enables developers to flexibly deploy robotic computing tasks based on varying computational and network conditions. We validate its effectiveness through experiments involving robotic arm control, 3D detection applications, and numerically simulated ROS2 tasks under different bandwidth conditions. The results demonstrate that the framework significantly improves response times for ROS2 applications, even under fluctuating computational loads and network constraints. The code for the framework ARDF can be found on GitHub <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup>.