The site is read-only. Please transition to use Robotics Stack Exchange
| ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
 | 1 | initial version |
I think the problem is running the driver twice in a single velodyne_nodelet_manager process.
We probably ought to incorporate the frame_id into the nodelet manager name name. Please open a github issue to get this resolved.
| | 2 | No.2 Revision |
I think the problem is running the driver twice in a single velodyne_nodelet_manager process.
We probably ought to incorporate the frame_id into the nodelet manager name name. Please open a github issue to get this resolved.
EDIT: to provide more detail, this command in velodyne_driver/launch/nodelet_manager.launch should be changed:
<node pkg="nodelet" type="nodelet" name="driver_nodelet"
args="load velodyne_driver/DriverNodelet velodyne_nodelet_manager" >
Instead of velodyne_nodelet_manager we should call it $(arg frame_id)_nodelet_manager, running the driver for each device in a separate process. Other nodelet loading to velodyne_nodelet_manager should similarly be changed. The node name may also need to be parameterized.
| | 3 | No.3 Revision |
I think the problem is running the driver twice in a single velodyne_nodelet_manager process.
We probably ought to incorporate the frame_id into the nodelet manager name name. Please open a github issue to get this resolved.
EDIT: to provide more detail, this command in velodyne_driver/launch/nodelet_manager.launch should be changed:
<node pkg="nodelet" type="nodelet" name="driver_nodelet"
args="load velodyne_driver/DriverNodelet velodyne_nodelet_manager" >
Instead of velodyne_nodelet_manager we should call it $(arg frame_id)_nodelet_manager, running the driver for each device in a separate process. Other nodelet loading to velodyne_nodelet_manager should similarly be changed. The node name may also need to be parameterized.
EDIT 2: there was a bug in the driver, now fixed in the master source. See ros-drivers/velodyne#108 for details.
Note that you need to configure the two devices to have different tf frames, IP addresses and UDP port numbers.
| | 4 | No.4 Revision |
I think the problem is running the driver twice in a single velodyne_nodelet_manager process.
We probably ought to incorporate the frame_id into the nodelet manager name name. Please open a github issue to get this resolved.
EDIT: to provide more detail, this command in velodyne_driver/launch/nodelet_manager.launch should be changed:
<node pkg="nodelet" type="nodelet" name="driver_nodelet"
args="load velodyne_driver/DriverNodelet velodyne_nodelet_manager" >
Instead of velodyne_nodelet_manager we should call it $(arg frame_id)_nodelet_manager, running the driver for each device in a separate process. Other nodelet loading to velodyne_nodelet_manager should similarly be changed. The node name may also need to be parameterized.
EDIT 2: there was a bug in the driver, provided launch files, now fixed in the master source. See ros-drivers/velodyne#108 for details.
Note that you need to configure the two devices to have different tf frames, IP addresses and UDP port numbers.
ROS Answers is licensed under Creative Commons Attribution 3.0 Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.