Tango updates the transform between area_description and start_of_service only every 0.1 s. This can cause some applications to work slower than expected.
For example, in the case of the ROS navigation stack, the control loop of the move_base package takes ~0.5 s. The reason is that it uses the base_local_planner which makes several calls to waitForTransform. Each call is blocking for between 0.1 to 0.2 s when waiting for the tf between area_description and start_of_service.
To avoid this, this PR replaces the use of tf by tf_static to publish the transform between area_description and start_of_service.
Tango updates the transform between area_description and start_of_service only every 0.1 s. This can cause some applications to work slower than expected.
For example, in the case of the ROS navigation stack, the control loop of the move_base package takes ~0.5 s. The reason is that it uses the base_local_planner which makes several calls to waitForTransform. Each call is blocking for between 0.1 to 0.2 s when waiting for the tf between area_description and start_of_service.
To avoid this, this PR replaces the use of tf by tf_static to publish the transform between area_description and start_of_service.