ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question

Best localization algorithms available in ROS2/ROS1.

asked 2021-09-13 07:33:01 -0500

root-robot gravatar image

I have used slam_toolbox and AMCL are these the only options. If we only have slam_toolbox and AMCL then which one is preferred strictly for localization(no mapping ) for large warehouses with a dynamic environment.

Thanks for your suggestions.

edit retag flag offensive close merge delete

2 Answers

Sort by » oldest newest most voted

answered 2021-09-14 01:37:55 -0500

So you say you aren’t looking for mapping, but remember that when you say “localize” you had have to localize yourself w.r.t something, a reference. I would guess what you mean is, you just want an odometry estimation, meaning that you want to recieve your localization information w.r.t your initial pose ? Is this what you mean ? If so, you are not limited to use laser scanner/lidar based slam methods. You can also use a camera based visual slam method such as RTABMap, OpenVSLAM or ORB-SLAM. I have used AMCL but haven’t used the other one, and as far as I remember you had need to provide a prebuilt map, and guessed initial pose to AMCL to start localization. In addition to both those methods there is google cartographer slam package,

edit flag offensive delete link more

answered 2022-12-04 15:13:57 -0500

updated 2022-12-04 15:33:13 -0500

If you have a changing or dynamic environment, SLAM_toolbox is the way to go! As it is demonstrated here: SLAM_toolbox performs way better than AMCL (achieving twice better accuracy). In the comparison, also Cartographer and GMCL are included!

Be aware that the comparison was made with a based map that only contains the permanent structures of the building. If you have a good quality map (e.g. from a Floor plan or architectonical model), you can use this tool to create a serialized .posegraph map and use it for localization with SLAM_toolbox!

edit flag offensive delete link more


How does is work outside? Like if I just wanted the path planning and obstacle detection side of it for planning and localization from gpa, IMU, encoders?

chased11 gravatar image chased11  ( 2022-12-05 02:17:35 -0500 )edit

The tool is mainly conceived to support localization in indoor GPS-denied environments, nonetheless, it could also work in outdoor environments. You still will required a map of the navigable space. Besides that, if you are outside and you have GPS you can fusion it with LIDAR and other data with a Kalman-Filter, as done, for example, here.

Miguel A. Vega Torres gravatar image Miguel A. Vega Torres  ( 2022-12-07 07:09:21 -0500 )edit

Perfect for my robot I think. Right now I am using the ROS navigation stack and am looking at the different laser-based packages. I have a 4cm accuracy GPS, at least 1 imu and might fuse 2, and encoder odom. I ask because since it uses localization there wouldn't be any issue when there are no walls, etc. to localize off of. Maybe I was looking at it wrong but definitely gonna try it. Thanks

chased11 gravatar image chased11  ( 2022-12-07 14:28:49 -0500 )edit

Question Tools



Asked: 2021-09-13 07:33:01 -0500

Seen: 568 times

Last updated: Dec 04 '22