LIONav (LiDAR-Inertial Odometry Navigation) is a high-performance autonomous navigation framework designed for GPS-denied environments such as underground mines, caves, industrial facilities, and large indoor spaces.
It integrates GPU-accelerated LiDAR–Inertial Odometry, real-time 3D mapping, ESDF-based planning, and intelligent Return-to-Launch (RTL) into a cohesive, orchestrated system suitable for aerial and ground robots.
- ouster-ros – Official ROS drivers for Ouster sensors (OS0, OS1, OS2, OSDome)
- FAST_LIO_GPU – LiDAR-IMU odometry leveraging GPU acceleration
- pose_graph_backend – Keyframe-based pose graph with scan-context place recognition and ICP/NDT loop closure
- drone_description – URDF and TF visualization for real-time inspection and debugging
- nvblox_fastlio_bridge – Converts Fast-LIO outputs into voxel-based TSDF/ESDF maps
- esdf_2_5d_planner – Hybrid planner combining: Local 3D ESDF for obstacle avoidance + Global 2D ESDF / occupancy for efficient long-range planning
- smart_rtl – Intelligent Return-to-Launch system combining: Pose-graph topology (guaranteed traversability) + ESDF-based local safety
- MAVLink_Bridge – Interface between LIONav and ArduPilot / PX4
- Orchestra – High-level behavior and mission orchestrator
MIT License © 2025 Ömer Mersin