The article substantiates the feasibility of using the Gazebo simulation environment as a digital scientific and experimental testbed for studying autonomous robotic systems. It is shown that the scientific potential of Gazebo lies not only in visualizing robot motion, but also in creating a reproducible, parameterized, and metrically controlled environment in which a world model, robot model, sensor system, algorithmic stack, experimental scenarios, research protocol, and data set can be explicitly defined. A methodological scheme of a Gazebo-based experiment is proposed as a formalized research loop that combines physical modeling, sensor data generation, ROS2 integration, scenario variation, and quantitative assessment of results. To strengthen the evidential value of simulation-based research, a system of metrics is developed, including localization error, map-building error, mission completion time, path length, number of collisions, near-collision events, replanning operations, mission success rate, and an integral mission risk indicator. Lemmas on the reproducibility of the simulation configuration, metric observability of results, and comparability of algorithmic configurations are formulated. On this basis, a theorem on the metric validity of a simulation experiment in Gazebo is proposed. A model-based approbation using the example of autonomous navigation of a mobile robot demonstrated a regular deterioration of performance indicators under increasing scenario complexity and degradation of sensor information. The obtained results confirm the possibility of using Gazebo as an intermediate methodological link between theoretical modeling, software implementation of algorithms, and subsequent physical validation of robotic systems.
Kutsaev et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: