This project explores the development of AI-based navigation systems for uncrewed aerial vehicles (UAVs) operating in challenging environments. With increasing demand for UAV applications in sectors like logistics, agriculture, and emergency response, the need for reliable navigation in complex conditions is critical. The research uses geospatial data and AI—particularly reinforcement learning—to simulate realistic environments with obstacles, variable weather, and GPS limitations. Initial results from simulation-based training indicate that UAVs can learn to navigate efficiently and safely, even under difficult conditions.

The integration of machine learning into UAV path planning demonstrates significant potential for improving autonomy, adaptability, and safety. However, challenges remain, including the need for extensive training data, computational demands, and the difficulty of modeling unpredictable real-world variables. These limitations point to the importance of transitioning from simulation to real-world testing in future phases.

Overall, the project lays a strong foundation for the future deployment of autonomous UAVs in imperfect environments. Continued development and validation will help realize systems that are not only more robust and efficient but also scalable across various real-world applications. By enhancing the reliability of UAV navigation, this research contributes to the broader goal of expanding the safe and effective use of uncrewed systems in everyday life.