Design and implementation of a containerized monitoring system for publish/subscribe networks

This work presents the design and implementation of a functional monitoring system for robotic applications. The monitoring system comprises components, tools, and processes to detect anomalies in the robot's behavior, alert users, and provide performance feedback. It also offers predictive maintenance by analyzing behavior over time and suggesting necessary repairs or parameter tunings. Developed using Robot Operating System (ROS), the monitoring system observes the behavior of a vehicle robot inside a simulation. Monitoring package includes tools, libraries, and processes to provide feedback on the robot's internal state (speed, actuator, and sensor status) and external conditions (environmental data, position, obstacles, other robots). It generates alarms, alerts, and notifications as needed, providing reports to enable robotic operations managers and operators to take appropriate action. The system is designed to be agile and efficient, capable of dynamically scaling according to domain-specific demands. Additionally, this work investigates the integration of runtime verification containers within publish/subscribe networks, which are crucial for applications like the Internet of Things (IoT) and robotics. Runtime verification techniques ensure the correctness of these highly dynamic and large-scale networks. By leveraging modern containerization technology, runtime verification tools are deployed to monitor message flow, ensuring real-time verification of network behavior. Each runtime verification container operates alongside other containerized participants, monitoring the system's integrity. This approach is benchmarked through several experiments and a real-world case study from the software-defined vehicle domain, demonstrating its performance.