MOLAorg/mola

MOLA – Modular Optimization framework for Localization and Mapping

What it is – MOLA is an open‑source C++ library (with ROS 2 integration) that provides a modular, high‑performance pipeline for robot localization and SLAM. It focuses on LiDAR‑based odometry, LiDAR‑Inertial odometry (LIO) and the creation of metric maps, all configurable through YAML files so you can swap sensors or algorithms without recompiling.

Key capabilities

• LiDAR Odometry (LO) & LiDAR‑Inertial Odometry (LIO) – real‑time pose estimation from 3‑D laser scans, optionally fused with IMU data.
• Geo‑referencing – combine low‑cost GNSS, LiDAR and IMU to produce RTK‑quality outdoor maps (< 1 cm accuracy) without a prior map.
• Configurable pipelines – YAML‑driven graphs let you enable/disable filters, map layers, or sensor inputs on the fly.
• Metric‑map ecosystem – the .mm format can be visualised, exported to LAS/PLY/TXT and processed with built‑in tools (filtering, georeferencing, etc.).
• ROS 2 ready – packages are built for Humble, Jazzy, Kilted, Lyrical and Rolling, but the core can also be used as a pure C++ library.

Typical use‑cases

• Autonomous ground or legged robots that need accurate pose tracking in GPS‑denied or GPS‑available environments.
• Survey‑grade mapping for construction, agriculture or infrastructure inspection where sub‑centimetre accuracy is required.
• Research projects that experiment with different sensor fusion strategies or map representations, thanks to the plug‑in architecture.

Ecosystem & documentation

• The repository hosts the core packages (mola, mola_bridge_ros2, mola_demos, various input adapters, etc.).
• Additional modules live in the broader MOLAorg GitHub organization.
• Full build instructions, API reference and demo videos are available at the official docs site: https://docs.mola-slam.org/latest/.
• Continuous‑integration pipelines provide binary builds for many ROS 2 distributions and platforms (amd64/arm64, Ubuntu, RHEL, Fedora).

Maturity

• Actively maintained with CI badges showing successful builds across multiple ROS 2 releases.
• Used in public demos (Oxford Spires, KITTI, legged‑robot datasets) and presented at ROSCon, indicating a level of production readiness.

License – GPL‑3.0, which permits free use and modification but requires derivative works to be open‑sourced under the same license.

---

In short, MOLA is a serious, production‑grade SLAM framework for robots that need flexible, high‑accuracy localization and mapping, especially when working with LiDAR, IMU and GNSS data.