Nicholas Hopf

Robot Operating System (ROS) applications are growing increasingly complex and distributed, incorporating multi-node architectures, real-time communication, and compute offloading to cloud or edge resources. Despite advances in containerization, large-scale robotics deployments often rely on ad hoc provisioning, updating, and scalability strategies. This paper’s key contribution is a module extension for an existing ROS-focused container framework that leverages Kubernetes’ capabilities—controlled rolling updates, ROS-adapted readiness checks, distributed workload management, and embedded observability—to bridge cloud-native practices with robotics development. The result is a unified tool that streamlines the entire ROS DevOps cycle. Experimental validation demonstrates stable operation with minimal message loss across diverse ROS communication patterns, including tests on an industrial mobile manipulator system. By simplifying orchestration and monitoring, this approach enables roboticists to concentrate on application logic rather than deployment and networking complexities.