Swarm Robotics (İLTER YAZILIM)
Teknofest 2022 Swarm Robotics finalist. As software team lead, I built the swarm control algorithm coordinating nano-drones over ROS.
Overview
We competed in the Teknofest 2022 Swarm Robotics Competition as İLTER YAZILIM, the team of the Science, Idea and Art community. I served as software team lead and built the swarm controller software.
Swarm Control Architecture
The system ran on a ROS "Master–Slave" topology: the referee (Master) PC published obstacle and target positions as ROS topics, while our team software (Slave) published our drones' poses and listened to rivals' and obstacles' positions to compute collision-free swarm motion. I controlled Crazyflie 2.0 nano-quadcopters over Crazyradio PA in Python, running takeoff/landing and position commands concurrently by assigning a thread per drone.
The Tasks
The competition had three tasks: obstacle avoidance, a cargo-carrying race against a rival swarm (collision avoidance via shared poses), and a mixed-swarm fire-extinguishing task together with a ground robot.
Outputs & Results
We placed 6th among the 10 finalist teams. I built swarm control software that coordinates multiple autonomous aerial vehicles in real time, collision-free — reaching the finals on our first swarm-robotics attempt.
ROS Master–Slave Architecture
Referee (Master · HAVELSAN)
Team Software (Slave · Python)
Swarm Hardware
Swarm Simulation
Tech Stack
Gallery
Building a robot/drone swarm system?
I can develop the ROS-based swarm control that coordinates multiple vehicles in real time, collision-free.
I'll build the same for you
Message the assistant — talk through whether I can build something like this, the approach and a ballpark, and get to a booking in minutes.