Simulation in real conditions of navigation and obstacle avoidance with PX4/Gazebo platform

• Published in: Personal and ubiquitous computing 2022-08, Vol.26 (4), p.1171-1191
• Main Authors: García, Jesús, Molina, Jose M.
• Format: Article
• Language: English
• Subjects: Collision avoidance; Computer Science; Flight control systems; Mobile Computing; Navigation; Obstacle avoidance; Original Article; Personal Computing; Sensors; Simulation models; Unmanned aerial vehicles; User Interfaces and Human Computer Interaction
• ISSN: 1617-4909; 1617-4917
• DOI: 10.1007/s00779-019-01356-4

In the future, UAVs should be a part of the IoT ecosystems. Integration of sensors onboard allows to enrich information stored in the cloud and, at the same time, to improve the capacities of UAVs. Developing new sensors and the integration in UAV architecture could improve control functions. Design of future UAV systems requires from advanced tools to analyze the system components and their interaction in real operational conditions. In this work, authors present an approach to integrate and evaluate a LIDAR sensor and the capacity for improving navigation and obstacle avoidance functions in simulated situations using a real UAV platform. It uses available software for mission definition and execution in UAVs based on PixHawk flight controller and peripherals. The proposed solution (a general method that could be used to integrate other kind of sensors) shows physical integration of the main types of sensors in UAV domain both for navigation and collision avoidance, and at the same time the use of powerful simulation models developed with Gazebo. Some illustrative results show the performance of this navigation and obstacle avoidance function using the simulated sensors and the control of the real UAV in realistic conditions.