3rd-party wrappers

For the sake of scientific comparison between different methods, we provide wrappers of other LiDAR odometry methods that mimic the interface of mola-lidar-odometry-cli, so exactly the same input datasets can be processed by different methods.

---

KISS-ICP

Wrapper for the work [VGM+23].

Repository: https://github.com/MOLAorg/mola_kiss_icp_wrapper

Compile instructions

Clone in your ROS 2 workspace:

mkdir -p ~/ros2_mola_ws/src/
cd ~/ros2_mola_ws/src/

git clone https://github.com/MOLAorg/mola_kiss_icp_wrapper.git --recursive

Install dependencies:

cd ~/ros2_mola_ws/
rosdep install --from-paths src --ignore-src -r -y

Compile:

cd ~/ros2_mola_ws/
colcon  build --symlink-install --cmake-args -DCMAKE_BUILD_TYPE=RelWithDebInfo

CLI reference

USAGE:

   mola-lidar-odometry-cli-kiss  [--input-mulran-seq <KAIST01>]
                                 [--kitti-correction-angle-deg <0.205
                                 [degrees]>] [--input-kitti-seq <00>]
                                 [--lidar-sensor-label <>] [--input-rosbag2
                                 <dataset.mcap>] [--input-rawlog
                                 <dataset.rawlog>] [--only-first-n <Number
                                 of dataset entries to run>] [--no-deskew]
                                 [--output-tum-path
                                 <output-trajectory.txt>] [--max-range
                                 <max-range>] [--min-range <min-range>]
                                 [--] [--version] [-h]


Where:

   --input-mulran-seq <KAIST01>
   INPUT DATASET: Use Mulran dataset sequence KAIST01|KAIST01|...

   --kitti-correction-angle-deg <0.205 [degrees]>
   Correction vertical angle offset (see Deschaud,2018)

   --input-kitti-seq <00>
   INPUT DATASET: Use KITTI dataset sequence number 00|01|...

   --lidar-sensor-label <>
   If provided, this supersedes the values in the 'lidar_sensor_labels'
   entry of the odometry pipeline, defining the sensorLabel/topic name to
   read LIDAR data from. It can be a regular expression (std::regex)

   --input-rosbag2 <dataset.mcap>
   INPUT DATASET: rosbag2. Input dataset in rosbag2 format (*.mcap)

   --input-rawlog <dataset.rawlog>
   INPUT DATASET: rawlog. Input dataset in rawlog format (*.rawlog)

   --only-first-n <Number of dataset entries to run>
   Run for the first N steps only (0=default, not used)

   --no-deskew
   Skip scan de-skew

   --output-tum-path <output-trajectory.txt>
   Save the estimated path as a TXT file using the TUM file format (see
   evo docs)

   --max-range <max-range>
   max-range parameter

   --min-range <min-range>
   min-range parameter

   --,  --ignore_rest
   Ignores the rest of the labeled arguments following this flag.

   --version
   Displays version information and exits.

   -h,  --help
   Displays usage information and exits.


   mola-lidar-odometry-cli-kiss

SiMpLE

Wrapper for the work [BPM24].

Repository: https://github.com/MOLAorg/mola_simple_wrapper

Compile instructions

Clone in your ROS 2 workspace:

mkdir -p ~/ros2_mola_ws/src/
cd ~/ros2_mola_ws/src/

git clone https://github.com/MOLAorg/mola_simple_wrapper.git --recursive

Install dependencies:

cd ~/ros2_mola_ws/
rosdep install --from-paths src --ignore-src -r -y

Compile:

cd ~/ros2_mola_ws/
colcon  build --symlink-install --cmake-args -DCMAKE_BUILD_TYPE=RelWithDebInfo

CLI reference

USAGE:

   mola-lidar-odometry-cli-simple  [--input-paris-luco] [--input-mulran-seq
                                 <KAIST01>] [--input-kitti360-seq <00>]
                                 [--kitti-correction-angle-deg <0.205
                                 [degrees]>] [--input-kitti-seq <00>]
                                 [--lidar-sensor-label <lidar1>]
                                 [--input-rosbag2 <dataset.mcap>]
                                 [--input-rawlog <dataset.rawlog>]
                                 [--only-first-n <Number of dataset
                                 entries to run>] [--output-tum-path
                                 <output-trajectory.txt>] -c
                                 <config.yaml> [--] [--version] [-h]


Where:

   --input-paris-luco
   INPUT DATASET: Use Paris Luco dataset (unique sequence=00)

   --input-mulran-seq <KAIST01>
   INPUT DATASET: Use Mulran dataset sequence KAIST01|KAIST01|...

   --input-kitti360-seq <00>
   INPUT DATASET: Use KITTI360 dataset sequence number 00|01|...

   --kitti-correction-angle-deg <0.205 [degrees]>
   Correction vertical angle offset (see Deschaud,2018)

   --input-kitti-seq <00>
   INPUT DATASET: Use KITTI dataset sequence number 00|01|...

   --lidar-sensor-label <lidar1>
   If provided, this supersedes the values in the 'lidar_sensor_labels'
   entry of the odometry pipeline, defining the sensorLabel/topic name to
   read LIDAR data from. It can be a regular expression (std::regex)

   --input-rosbag2 <dataset.mcap>
   INPUT DATASET: rosbag2. Input dataset in rosbag2 format (*.mcap)

   --input-rawlog <dataset.rawlog>
   INPUT DATASET: rawlog. Input dataset in rawlog format (*.rawlog)

   --only-first-n <Number of dataset entries to run>
   Run for the first N steps only (0=default, not used)

   --output-tum-path <output-trajectory.txt>
   Save the estimated path as a TXT file using the TUM file format (see
   evo docs)

   -c <config.yaml>,  --config-file <config.yaml>
   (required)  Simple config file

   --,  --ignore_rest
   Ignores the rest of the labeled arguments following this flag.

   --version
   Displays version information and exits.

   -h,  --help
   Displays usage information and exits.


   mola-lidar-odometry-cli-simple

DLIO

Wrapper for the work [CNL23].

Repository: https://github.com/MOLAorg/mola_dlio_wrapper

Unlike the filter-based methods above, DLIO uses no filter and no factor graph: the pose comes from GICP scan-to-submap registration, and the IMU is fused by a nonlinear geometric observer with constant gains. Every point is deskewed against a continuous IMU-integrated trajectory rather than a per-scan linear interpolation, and the local map is a submap selected from keyframes by a hull plus kNN search rather than a voxel or octree structure.

It provides two entry points:

• mola::DlioOdometry, an online module loadable from a mola-cli launch YAML (type: mola::DlioOdometry), which publishes localization and submap updates so the MOLA GUI draws the trajectory and the growing map live.

• mola-dlio-cli, an offline tool that waits for each observation to finish before feeding the next, so no scan is ever dropped. This is the one to use for comparisons: see GUI or offline CLI: which one, and why.

Note

This wrapper currently targets Oxford Spires, read through the generic mola::Rosbag2Dataset input. It does not cover the full set of dataset sources that the KISS-ICP and SiMpLE wrappers accept.

Compile instructions

Clone in your ROS 2 workspace:

mkdir -p ~/ros2_mola_ws/src/
cd ~/ros2_mola_ws/src/

git clone https://github.com/MOLAorg/mola_dlio_wrapper.git --recursive

Install dependencies and compile:

cd ~/ros2_mola_ws/
rosdep install --from-paths src --ignore-src -r -y
colcon build --symlink-install --packages-select mola_dlio_wrapper \
  --cmake-args -DCMAKE_BUILD_TYPE=RelWithDebInfo
source install/setup.bash

GLIM

Wrapper for the work [KYOB24].

Repository: https://github.com/MOLAorg/mola_glim_wrapper

Upstream GLIM is a full SLAM system: an odometry front end, a sub-mapping stage, and a global-mapping back end with loop closure. This wrapper instantiates the odometry stage only (glim::OdometryEstimationCPU, a fixed-lag factor-graph estimator over GICP/VGICP scan-to-model factors and preintegrated IMU factors).

That is deliberate. The wrapper exists for cross-method odometry comparison, and a loop-closed, globally optimized trajectory is not comparable with the odometry-only results the other wrappers on this page produce.

It provides:

• mola::GlimOdometry, an online module loadable from a mola-cli launch YAML.

• mola-glim-cli, an offline, loss-free CLI over the MOLA dataset sources: KITTI, ROS 1 bags, ROS 2 bags, MulRan and rawlog.

• Pipeline YAMLs for Oxford Spires, KITTI, BotanicGarden and Newer College.

Compile instructions

Everything GLIM needs is vendored in the repository, so the only prerequisites are system packages: GTSAM >= 4.2 with gtsam_unstable (ros-$ROS_DISTRO-gtsam on ROS distributions), Eigen 3, Boost (graph, filesystem, serialization), spdlog, OpenMP, plus MRPT and the MOLA core packages.

mkdir -p ~/ros2_mola_ws/src/
cd ~/ros2_mola_ws/src/

git clone https://github.com/MOLAorg/mola_glim_wrapper.git --recursive

cd ~/ros2_mola_ws/
rosdep install --from-paths src --ignore-src -r -y
colcon build --symlink-install --packages-select mola_glim_wrapper \
  --cmake-args -DCMAKE_BUILD_TYPE=RelWithDebInfo
source install/setup.bash

Example: run on a KITTI sequence

KITTI_BASE_DIR=/path/to/kitti mola-glim-cli \
  -c $(ros2 pkg prefix mola_glim_wrapper)/share/mola_glim_wrapper/pipelines/glim-kitti.yaml \
  --input-kitti-seq 04 \
  --output-tum-path /tmp/glim_kitti04.tum