From Google Open Source

Damon Kohler, Wolfgang Hess, Holger Rapp

Heart of the machine compiles

Participation: Li Yazhou

Today, Google is open sourcing Cartographer, a library of SLAM technology needed for self-driving cars, robots and other platforms.

  • Open source web site: https://github.com/googlecartographer

Today, we’re excited to announce open source Cartographer, a 2D and 3D SLAM (Simultaneous localization and Mapping) library supported by ROS systems.

SLAM algorithms combine data from multiple sensors (for example, LIDAR, IMU, and camera) to synchronously calculate the sensor’s location and map the environment around the sensor. For example, use this method to draw a floor plan of a living room:

  • Stand in the middle of the house with a laser rangefinder and draw an X on a piece of paper.

  • Measure the distance between your position and any wall;

  • Draw a line where the wall is and write the distance between X (where you are) and the wall.

  • Measure the distance between your position and another wall and draw a straight line representing the wall.

  • Now, move to another part of the house;

  • Since the walls do not move, you can measure the distance between the two walls and determine your position.



SLAM is an important component of many autonomous driving platforms, such as autonomous vehicles, automated forklifts in warehouses, floor sweeping robots, UAVs, and more.

Cartographer builds globally consistent maps in real time on sensor configurations common to industry and academia. As shown in the following video:

Cartographer 2D: Real-time Loop Closure in 2D LIDAR SLAM real-time Loop Closure in 2D LIDAR SLAM

Cartographer has been used on multiple ROS-enabled robot platforms due to the integration of ROS and support from external contributors:

  • Toyota HSR

  • TurtleBots

  • PR2

  • Revo LDS

At Google, Cartographer has been used in a wide range of applications, from museums and transportation hubs to visualizations of famous buildings.

We recognize the value of high-quality data sets to the research community, which is why in collaboration with the German Museum, the largest science museum in the world, Open source Lidar and IMU data collected during the development and testing of Cartographer for 3 years using the 2D and 3D Drawing Backpack platforms.

Cartographer is currently particularly focused on Lidar SLAM, and through community contributions and continued development, we hope to add more sensor and platform support and add more new features such as Lifelong Mapping and localizing in pre-existing maps.

© This article is compiled by Machine Heart, please contact this official account for authorization.

✄ — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —

Join Heart of the Machine (full-time reporter/intern) : [email protected]

Contribute or seek coverage: [email protected]

Advertising & Business partnerships: [email protected]