Bio-Inspired AUV — ROS2 Control, IMU Fusion, and Camera–LiDAR Obstacle Avoidance

TL;DR: A fin-driven AUV with PID control and multi-sensor fusion (IMU + vision + LiDAR). Containerized ROS2 stack for repeatable bring-up and deployment.

Highlights

• Stability: Improved yaw/pitch/roll stability with tuned PID and IMU fusion (EKF)
• Perception: Camera–LiDAR fusion for obstacle avoidance
• Deployment: Dockerized nodes for reproducibility on embedded compute
• Result (project): ~30% better trajectory stability and improved energy efficiency

Repo Layout

• src/ — ROS2 packages (control, fusion, perception)
• launch/ — launch files (sim vs hardware)
• config/ — PID gains, EKF params, topics
• scripts/ — calibration, data logging, analysis
• docker/ — Dockerfile & compose
• sim/ — simulation worlds/models (optional)
• media/ — demos (GIF/MP4)
• docs/ — design notes / report

Quick Start (dev)

# Python deps not coming from ROS
python -m venv .venv
# Windows:
.\.venv\Scripts\activate
pip install -r requirements.txt

# ROS2 workspace (example)
# colcon build
# source install/setup.bash
# ros2 launch auv_bringup sim.launch.py