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

TF broadcaster on robot , where to start?

asked 2017-11-10 00:09:45 -0500

Hey , i am new to ROS . Currently i am working on this robotimage description(http://)(http://)(http://).


  1. sabertooth 2x32 motor driver
  2. 18aH 12v battery
  3. raspberry pi
  4. Logitech USB camera
  5. Line laser
  6. 2X 250watt ebike motors

The current configuration doesn't have a encoders . This is a big issue for determining the odom .

The bot is currently partially ROS enabled (Skid steer ) . I want to configure the bot with TF such that the camera and wheels are in the right perspective .

I am not able to figure out where to start , should i write a URDF/XACRO description of the robot and add the camera to it ?

If anyone knows some simple tutorials please point me towards that .

edit retag flag offensive close merge delete


I've given you enough karma to attach images. Please edit your question to add the image in.

gvdhoorn gravatar image gvdhoorn  ( 2017-11-10 00:52:45 -0500 )edit

1 Answer

Sort by ยป oldest newest most voted

answered 2017-11-10 00:48:02 -0500

gvdhoorn gravatar image

updated 2017-11-10 00:53:29 -0500

We have somewhat of an expression in my lab:

If you're using ROS, but don't have a urdf, you're doing it wrong

that is of course grossly generalising, but there is some truth there.

Whether or not you have encoders, start creating a urdf modelling the structure of your robot.

Worry about adding other things later (such as "TF broadcasters": you'll know when you need them, right now, you don't yet).

Also note: encoders (ie: wheels) are not the only source of odometry. Visual odometry and IMUs are also good sources.

edit flag offensive delete link more


Thanks , you are right . I should leave everything and first focus on creating the URDF model . I am already on it .

chris_sunny gravatar image chris_sunny  ( 2017-11-10 01:35:53 -0500 )edit

Question Tools


Asked: 2017-11-10 00:09:45 -0500

Seen: 183 times

Last updated: Nov 10 '17