Optimized flocking of autonomous drones in confined environments.

Area of Science:

• Robotics
• Artificial Intelligence
• Control Systems

Background:

• Existing flocking models for aerial robots often fail on real hardware due to neglecting constraints.
• Real-world multi-robot systems face challenges like limited communication, delays, and obstacles.

Purpose of the Study:

• To develop and validate a robust flocking model for autonomous drones operating in confined environments.
• To address limitations of current models by incorporating realistic constraints and an optimization framework.

Main Methods:

• Proposed a novel flocking model for drones integrated with an evolutionary optimization framework.
• Utilized carefully selected order parameters and fitness functions for model development.
• Conducted numerical simulations and field experiments with a swarm of 30 drones.

Main Results:

• The flocking model demonstrated stable collective motion in large drone swarms, even at high velocities and near obstacles.
• The system exhibited coherent and realistic collective motion patterns under perturbed conditions.
• Successfully validated the model on a real-world outdoor system of 30 self-organized drones without central control.

Conclusions:

• The developed flocking model effectively handles realistic constraints for autonomous drone swarms.
• This work represents the largest reported outdoor aerial system exhibiting flocking with collision and object avoidance.
• The approach enables more efficient task management for large drone swarms in diverse applications.