Optimizing Route Planning in Diverse Landscapes: Integrating SLAM and RRT in Autonomous Drone Deployment.

This thesis focuses on the implementation of improved route planning and terrain mapping in different types of structured landscapes by using advanced Simultaneous Localization and Mapping (SLAM) methods and Rapidly-exploring Random Tree (RRT*) algorithm in autonomous drone deployment, focusing on software development and simulation to improve tracking capabilities. This thesis is based on a single project that is a collaborative work of postgraduate students. The project aims to develop a drone-based telecommunications network that serves as a basis for exploration and monitoring in studied or designated areas. This research is based on the establishment of a system that integrates SLAM, which provides a quick and accurate map of complex environments. This is important for the correct drones' work and the best results. Meanwhile, the inclusion of an RRT algorithm enables us to raise the system's efficiency and accuracy in planning routes for drones as they navigate intricate urban and non-urban spaces. The project is exclusively based on software simulation, using tools like AirSim and Unreal Engine, which allow the creation of an environment to test how well different drones work, from a single area to specific coordinates, ignoring external conditions like weather and drone battery life. The use of these mapping techniques and this trajectory planning algorithm enables safe navigation and an understanding of the environment that allows drones to function properly to perform their tasks. The results obtained, and the methods applied in the thesis, hope to introduce efficiency and productivity in the planning of drone deployments in all structured environments. This path would open the way to new applications in areas beyond urban infrastructure.