A computationally efficient and robust LiDAR-inertial odometry (LIO) package using a tightly-coupled iterated Kalman filter.
FAST-LIO is an open-source LiDAR-inertial odometry package that provides real-time, robust state estimation for robots and autonomous vehicles. It solves the problem of accurate localization and mapping in dynamic, noisy, or feature-deficient environments by tightly fusing LiDAR and IMU data. The system is designed to be computationally efficient, enabling high-frequency operation on embedded platforms.
Robotics researchers and engineers developing autonomous drones, ground vehicles, or mobile robots requiring reliable, real-time localization and mapping. It is particularly suited for those working with LiDAR-inertial sensor suites.
Developers choose FAST-LIO for its high computational efficiency, robustness in degenerate environments, and direct odometry approach that works with raw LiDAR points. Its support for a wide range of LiDAR hardware and ability to run on ARM-based systems make it a versatile and practical solution.
A computationally efficient and robust LiDAR-inertial odometry (LIO) package
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Eliminates mandatory feature extraction by working with raw LiDAR points, improving accuracy and supporting diverse LiDAR types like Velodyne and Livox.
Uses incremental ikd-Tree mapping to achieve over 100Hz LiDAR rates, enabling real-time performance on embedded platforms like Raspberry Pi.
Tightly-coupled iterated Kalman filter integrates LiDAR and IMU data in a single loop, ensuring reliability in fast-motion or cluttered environments.
Compatible with spinning LiDARs (e.g., Ouster), solid-state LiDARs (e.g., Livox Avia), and ARM-based systems, as highlighted in the README.
Requires ROS, specific drivers like livox_ros_driver, and manual extrinsic calibration, making setup tedious and error-prone for newcomers.
Performance degrades without accurate hardware time sync between LiDAR and IMU; software sync is less reliable, as warned in the README.
Primarily focuses on LiDAR-IMU; adding cameras requires separate extensions like FAST-LIVO, not core to the package.