AeroTerrainBot is an advanced robotics research project focusing on Multi-Modal Mobility. Inspired by the M4 (Multi-Modal Mobility Morphobot), this platform seamlessly transforms between an All-Terrain Wheeled Vehicle (Tank Mode) and a Quadrotor Drone (Flight Mode).
Designed for high-versatility missions, it solves the "last-mile" problem in complex environments where traditional ground or aerial robots alone would fail.
- Kinematic Transformation: Utilizes high-torque servos and linear actuators to reconfigure the chassis geometry in real-time.
- Dual-Mode Propulsion:
- Ground: 4x Skid-Steer Omni-wheels driven by independent velocity controllers.
- Aerial: 4x Internal brushless rotors exposed by rotating the wheel-booms 90° vertically.
- Physics-Centric Simulation: Full simulation in Gazebo Harmonic, featuring contact-friction dynamics, IMU, Lidar, and Camera sensor integration.
- Modular ros2_control Framework: A sophisticated hardware abstraction layer that allows the exact same control logic to run in simulation (
GazeboSimSystem) and on physical hardware (Teensy 4.0).
| Ground Drive (Tank Mode) | Aerial Propulsion (Flight Mode) |
|---|---|
 |
 |
| Wheels down, booms horizontal for ground traversal. | Booms rotated 90° up; wheels serve as prop-guards. |
- Middleware: ROS 2 Jazzy Jalisco (Ubuntu 24.04)
- Simulator: Gazebo Harmonic (GZ Sim)
- Control:
ros2_control,joint_state_broadcaster,diff_drive_controller - Modelling: XACRO / URDF with nested macros for modular boom/wheel assemblies
- DevOps: Docker, Docker Compose for reproducible builds
The architecture utilizes a centralized controller_manager to orchestrate three specialized controllers:
- Tank Drive Controller: A
diff_drive_controllervariant for skid-steer ground movement. - Morph Controller: A
joint_trajectory_controllermanaging the 2 servos and 2 linear actuators for state transitions. - Rotor Controller: A
velocity_controllergroup for independent rotor RPM management during flight.
This repository is fully containerized for easy setup.
- Docker & Docker Compose
- VcXsrv / XLaunch (For Windows users to view the GUI)
# Start the container
docker compose up -d
# Build the workspace (inside the container)
docker exec -it aeroterrabot-aeroterrabot_dev-1 bash -c "source /opt/ros/jazzy/setup.bash && cd /workspace/aeroterrabot_ws && colcon build --symlink-install"Launch the Gazebo world, robot spawner, and ROS-GZ bridge:
docker exec -it aeroterrabot-aeroterrabot_dev-1 bash -c "source /opt/ros/jazzy/setup.bash && source /workspace/aeroterrabot_ws/install_ros/setup.bash && ros2 launch aeroterrabot_gazebo gazebo.launch.py"Drive the robot using your keyboard:
docker exec -it aeroterrabot-aeroterrabot_dev-1 bash -c "source /opt/ros/jazzy/setup.bash && ros2 run teleop_twist_keyboard teleop_twist_keyboard --ros-args --remap cmd_vel:=/tank_drive_controller/cmd_vel"- Simulator Synchronization: Resolved clock-desynchronization issues between the ROS 2 node and the Gazebo physics engine by implementing synchronized
use_sim_timeparameters. - Hardware Abstraction: Designed a unified URDF that dynamically toggles between serial-based hardware interfaces and Gazebo system plugins based on launch arguments.
- Physics Tuning: Optimized friction coefficients and chassis weight distribution to prevent high-centering during high-speed ground maneuvers.
- Autonomous Navigation: Integrating Nav2 stack for SLAM-based path planning.
- Flight Controller Integration: Implementation of PX4/ArduPilot SITL for aerial stability.
- Computer Vision: Deployment of YOLOv8 nodes for obstacle detection via the onboard simulated camera.
Developed by Mohammed Faraz. Building the future of hybrid mobility.