Revision history [back]

Alright, so this is the issue. The fixed_frame is the frame in which all readings are put together by the laser_assembler. So if you move your robot, collect data, and put them together into a single(!) Point cloud, the fixed_frame needs to be fixed in the map. I.e. you could either use the odometry or the localization frame. But note that in both cases you'll end up with errors.

The typical use case for the laser assembler is to e.g. tilt a laser on the robot platform (while this is not moving) and assemble the scans into a 3D sweep of the Environment.