Chang-Tien Lu, professor of computer science and associate director of the Sanghani Center for Artificial Intelligence and Data Analytics, is leading a team of Ph.D. students -- Ganyu Zhou, Siji Chen, Linhan Wang and Chen Dai -- on a project that uses artificial intelligence to help groups of drones coordinate safely and efficiently without a central controller. 

They designed Spatio-Temporal Graph Neural Network (STGNN), which allows each drone to make an intelligent decision using only neighboring drones and past motion patterns, and similar to birds flying in formation. (See video below.)

The system supports multiple formations, including clusters, triangular groups, and structured moving swarms while following leaders and avoiding obstacles.

The technology has been validated through a real-world Crazyflie drone swarm demonstration, where multiple palm-sized drones fly indoors, maintain coordinated formations, adjust positions in real time, and navigate as a unified group.

The researchers believe these advances could support future applications in disaster response, logistics, infrastructure inspection, defense, and environmental monitoring, where many autonomous systems must work together safely and efficiently.

Virginia Tech audio

Small drones flying in formation

29  Jul  2026