5G-enabled robots: Differentiated connectivity forvarying mission requirements through dynamic QoS

The fifth generation (5G) cellular network technology has matured and is increasingly utilized in many industrial robotics applications. Various sectors seek to harness the advanced communication performance and deploy time-critical robotic applications in a connected manner, often hosting the execution of computational intensive time-critical components in the edge cloud. Robust deployment of cellular enabled robots that rely on the network performance demands the utilization of quality of service (QoS) solutions to respect the real-time requirements of critical applications. This paper proposes a method of harnessing the 5G QoS features in a dynamic fashion to retain the time-critical requirements of the edge-offloaded robotics applications. The paper emphasizes the dynamic selection of network resources considering the continuously changing communication requirements of such applications based on the underlying evolving mission and highlights the importance of deploying offloaded robotics applications to be communication aware --- towards the era of co-design. Further, the dynamic nature of the background network traffic is examined and robot scalability with varying mission priorities is analyzed and discussed. A novel modeling approach coupling the time-critical performance of 5G-enabled robotics and the dynamic QoS selection is presented and utilized in the selection of the appropriate QoS profile. The lack of real-life evaluation of complete similar solutions is tackled by extensive experimental evaluation utilizing a real-life 5G stand alone (SA) network and a quadruped robot. The obtained results demonstrate the importance of such synergistic solutions.