Teleoperation of Robots via Simulated Networks: Evaluating Latency Impact on Start/Stop Control in ROS 2

In disaster response scenarios, mobile robots reduce safety and operational risks for human operators while enabling exploration of hazardous environments. Their effectiveness, however, strongly depends on the reliability and latency of the communication infrastructure connecting the operator to the robot. This thesis investigates the feasibility of teleoperating a ROSbot XL within a ROS 2 environment under varying network constraints. A Gazebo-based simulation integrated with ROS 2 Jazzy is used, where the robot performs a standardized start/stop task in response to obstacles, with latency and stopping distance serving as key performance indicators. Network degradation is emulated through artificial latency and packet loss injection via netns, reflecting conditions analogous to 5G, Wi-Fi, and satellite connectivity. MATLAB is employed for data analysis of reaction times and control stability. The results highlight the trade-offs between low-latency terrestrial networks, which enable precise and reliable teleoperation, and high-latency satellite links, which compromise responsiveness and safety. To complement the simulation, extensive experiments were also conducted on a physical ROSbot XL. These real-world tests were crucial to validate the fidelity of the simulation results, as they account for sensor imperfections, actuator delays, and the inherent variability of physical environments. This work provides a reproducible methodology for assessing the resilience of robotic teleoperation systems in challenging communication environments.