Heterogeneous Robot Coordination
Project Overview
Motivation
The primary goal of this project was to develop a robust system for coordinating heterogeneous robots, including drones and quadrupeds, to perform tasks such as formation control, obstacle avoidance, and synchronized movement. The project aimed to explore the integration of ROS2, Vicon systems and object detection and avoidance to enable seamless collaboration between different robotic platforms.
What Was Done
ROS2 Installation
Installed ROS2 on Raspberry Pi 4B and Raspberry Pi 2W via docker.
System Integration
Integrated Vicon systems, and ZED cameras for precise positioning and navigation.
Software Development
Developed ROS packages to control the agents, object detection, and path planning.
Testing & Integration
Conducted over 130 tests, including formation control and obstacle avoidance. Used ROS-Bridge for seamless communication between ROS1 and ROS2.
What was Learned
Robot Synchronization
Synchronization between heterogeneous robots is critical for effective coordination.
Communication Challenges
Communication reliability is a significant challenge in multi-robot systems, especially in bandwidth-constrained environments.
Sensor Fusion
Object detection models must adapt to varying lighting conditions and environments. Just Camera is not enough in outdoor conditions - it needs to be complemented with LIDAR or other sensors for autonomous navigation.
What Was Achieved
- Successfully demonstrated drone and quadruped coordination
- Developed robust ROS2 package for object detection
- Implemented obstacle avoidance and path planning
- Published results in academic papers