Collision-free swarm take-off based on trajectory analysis and UAV grouping

In recent years, the adoption of unmanned aerial vehicles (UAVs) has widely spread to different sectors worldwide. Technological advances in this field have made it possible to coordinate the flight of these aircraft so as to conform a swarm. A UAV swarm is defined as a group of UAVs working collaboratively to carry out more complex missions or perform tasks more efficiently. Common applications of these swarms include rescue missions, precision agriculture, and border control, among others. However, there are still certain problems that prevent us from ensuring the success of their mission, especially as the number of drones in a swarm increases. In this paper, we specifically address the problem of a swarm take-off by optimizing the total time involved, while guaranteeing the safety of the UAVs during the take-off stage. To this end, we propose a new approach that combines a collision detection algorithm based on trajectory analysis with a batch generation mechanism that we use in order to determine the take-off sequence. Experiments show that our algorithm offers an efficient solution, managing to improve the performance of existing take-off techniques.