Cloud-based remote-controlled robot system: A Kubernetes-Orchestrated approach to high availability

Abstract

Cloud-based remote-controlled robotic systems are increasingly being adopted in industrial automation due to their flexibility and scalability. However, ensuring high availability and real-time responsiveness remains a significant challenge, particularly in applications sensitive to service downtime. In this paper, we present the design and implementation of a Kubernetes-based architecture for remote robot control, emphasizing high availability and fault tolerance. The robot control engine is containerized and deployed using a dual-pod setup consisting of an active and a standby pod. A load balancer is employed to seamlessly route control commands to the active pod while monitoring system health. Upon failure of the active pod, the standby pod automatically takes over, ensuring continuous operation. The system is evaluated using a self-balancing robot that requires uninterrupted real-time feedback. Experimental results validate the proposed architecture's effectiveness in minimizing downtime and demonstrate its suitability for high-availability robotic control in cloud-based environments.