Cartographer

Cartographer is a standalone C++17 library developed by Google that provides real-time simultaneous localization and mapping (SLAM) capabilities for robotics and autonomous systems. It processes sensor data from LiDAR, IMU, and odometry to build 2D and 3D maps while tracking the robot's position.

The system uses a submap-based approach where local SLAM builds submaps from sensor data, and global SLAM detects loop closures to optimize the overall map consistency. This architecture enables real-time operation on commodity hardware while maintaining accuracy for large-scale environments.

Cartographer integrates with ROS (Robot Operating System) through a separate package, making it accessible to the broader robotics community. The library supports pure localization mode for navigating known maps and multi-trajectory SLAM for collaborative mapping scenarios.

Use cases

• Autonomous mobile robots (AMRs) navigating warehouse and factory floors
• Self-driving vehicles building high-definition maps for navigation
• Indoor mapping for facility management and digital twin creation
• Drone navigation for inspection and surveying applications

Limitations

• Requires 64-bit architecture; 32-bit builds have known memory alignment issues that cause crashes
• Minimum 16 GB RAM recommended for processing large point clouds
• Tuning parameters for specific sensor configurations requires expertise and iterative testing
• Last significant update was January 2024; maintenance activity has slowed
• ROS2 support is community-maintained rather than officially supported
• Limited built-in tools for map post-processing and export compared to commercial alternatives