MAD-ICP

Software
Navigation and mapping
Library

MAD-ICP is a LiDAR odometry pipeline that estimates a sensor's motion from point clouds, aiming for robust, real-time operation. It was accepted to RA-L 2024.

Made by Robots Vision and Perception Group, Italy (checked on github.com/rvp-group (opens in a new tab)).

Source on GitHub: github.com/rvp-group/mad-icp (opens in a new tab). Facts read at commit cb47d83 (opens in a new tab).

Licenses

Code: BSD-3-Clause (opens in a new tab)
Read by Robotheca from LICENSE (opens in a new tab) (BSD-3-Clause) at commit cb47d83 (opens in a new tab), on 16 Oct 2026.
Commercial use: Yes

What it does

Runs LiDAR odometry on Rosbag1, Rosbag2 and KITTI binary data via a Python launcher.
Writes the odometry estimate in KITTI format as row-major homogeneous matrices.
Dataset configurations cover sensor characteristics and extrinsics for many datasets.
Internal parameters are set in a default configuration file and can be replaced with a custom one.
Parameters such as number of keyframes and cores can be tuned for speed, and it is described as anytime realtime.
Includes a MAD-tree for nearest neighbor search and a registration tool between two point clouds.

Written by Robotheca from the project's README (opens in a new tab) at commit cb47d83.

Getting started

Difficulty: Intermediate
How to start: The project's README (opens in a new tab)

Maintenance

Status: Maintained: a commit in the last 12 months
Newest commit: 20 Feb 2026 on main
Latest release: 0.0.10 (16 Jun 2025)

Checked on GitHub on 16 Oct 2026.

Robots that can run it

No robot on Robotheca is linked to it yet. Links come from a robot's robot.yaml, from its dependency files read as text, or from Robotheca's editors.

Robotheca lists software as a reference: nothing is hosted or run here, and every fact links to where it was read.