Leader–Wingman Swarm Robots
A leader–follower robot platform: the leader navigates on its own using LiDAR, an IMU and a camera while follower robots hold formation behind it. Built from scratch on ROS 2, with computer-vision pose estimation and sensor fusion tying perception to control.
Team
Hyunjun An, Youngmin Jeon, Siwoong Lee, Jung Shim, Brian Cho, Eben Lee, Jinwoo Baek — Electrical & Computer Engineering, Mechanical Science & Engineering, and Physics.
Motivation
Autonomous systems are increasingly sent into places that are unsafe or inaccessible for people. This project builds a small version of that idea: one robot leads and senses the environment while others follow in formation, extending an operator’s reach past what they can see themselves.
What we built
- A leader robot on a Raspberry Pi 5 with 2D LiDAR, an IMU and a camera, running ROS 2 so perception, estimation and control live in separate nodes.
- Follower robots on ESP32 with an IMU and an ultrasonic sensor, holding position relative to the leader.
- Power distribution and motor-driver electronics designed and integrated from scratch.
- Computer vision with ArUco markers and OpenCV to measure how far and at what angle each follower sits from the leader.
- Kalman filtering so the estimated state settles quickly despite noisy sensor readings, and recovers after an abrupt change.
Next steps
Formation control with three robots, LiDAR–camera calibration, SLAM-based mapping, and real-time object detection.
Shown at
Engineering Open House 2026, University of Illinois Urbana-Champaign.
Gallery
