MOLAorg/mola

A Modular Optimization framework for Localization and mApping (MOLA)

What it solves

MOLA provides a modular and high-performance framework for robotics localization and mapping (SLAM). It addresses the need for flexible, configurable pipelines that can handle various sensor combinations (like LiDAR and IMU) without requiring constant code recompilation.

How it works

It is built as a C++ and ROS 2 framework that allows users to define pipelines via YAML configuration files. This modularity enables the integration of different sensors, filters, and map layers. The system supports various odometry methods, including LiDAR Odometry (LO) and LiDAR-Inertial Odometry (LIO), and can achieve high-precision georeferencing using GNSS, LiDAR, and IMU fusion.

Who it’s for

This framework is designed for robotics developers and researchers working on autonomous navigation, high-accuracy mapping, and sensor fusion in outdoor or indoor environments.

Highlights

  • Configurable Pipelines: Change sensors and filters via YAML without recompiling.
  • Multi-Sensor Fusion: Supports LiDAR, IMU, and GNSS for RTK-quality georeferencing.
  • Flexible Accuracy: Scales from fast real-time navigation to survey-grade mapping (<1 cm).
  • Broad Compatibility: Fully ROS 2 ready (Humble, Jazzy, Kilted, Rolling) and usable as a standalone C++ library.
  • Metric Map Ecosystem: Includes tools for viewing, filtering, and exporting maps in LAS, PLY, and TXT formats.