A recommendation system for requirements tuning of BVLoS drones

The purposes for which Unmanned Aerial Vehicles (UAVs) are used today are innumerable and their role is continuously expanding. In particular, BVLoS (Beyond Visual Line of Sight) drones, that enable large-scale and long-distance missions, pose complex challenges such as managing dynamic obstacles, connectivity issues and energy constraints. Drone performance heavily depends on interdependent parameters such as altitude, speed and battery usage where tuning one affects others, but such interdependencies have not been adequately addressed in the existing drone software. This paper introduces a generic approach for BVLoS mission planning through an interactive recommendation system that considers both parameter dependencies and operational requirements. The framework includes parameter taxonomy creation, requirement classification as functional and non-functional ones, dependency graph modeling considering both static and dynamic parameters, and an optimization algorithm to generate tuning recommendations. Extensive simulations and implementation results demonstrate improved mission efficiency, lower computational complexity, and higher reliability, providing a structured, scalable solution for BVLoS operations under diverse environmental and mission-specific scenarios.