Open-Source Software-in-the-Loop Strategies for Realistic UAV Monitoring Applications

The rise in UAV capabilities has significantly boosted aerial monitoring activities in recent years. This surge in un-manned aerial vehicles has increased attention towards monitoring tasks in vital areas, such as urban spaces for security, surveillance or Internet of the Things applications. In this sense, we propose an open-source software-in-the-loop approach that integrates the complete navigation experience, accounting for potential wind-related effects, and providing realistic 3D modeling of scenarios encompassing open spaces in city landscapes. Our work introduces a wind perturbation estimation strategy, which models the wind effect as an external force along the Z-axis based on altitude variations from real flights. Validation was conducted using real data flights with a quadrotor aiding the navigation with Ardupilot and QgroundControl as a ground station. To complement the experiments we include a small test with a fixed-wing plane to asses the software-in-the-loop accuracy on larger trajectories. Results demonstrate successful tracking of altitude variations, yielding errors below 10% relative to the home position in experiments with average trajectory lengths of 150 m in two scenarios in Quito. The software-in-the-loop simulation aimed to replicate wind conditions specific to the date and time of real data acquisition, providing a platform to emulate possible improvements and assess UAV performance before deployment in actual missions. This research not only contributes to the optimization of monitoring capabilities but also serves as a valuable tool for assessing UAV missions in diverse and dynamic environments.