Are there any additional method to avoid dust, snow and rain? I think vertical ray is really reasonable method for object detection, but really sensitive to noise.
It would definitely be possible, however we never really tested something like this. We most of the time work with the 3D sensors like e.g. Velodyne. For your use case I would suggest checking out navigation.ros.org/.
We have not tested severe slopes on the road. That is the use case in which you would need to use also other sensors (e.g. IMU) and HD map information.
It is compulsory to have Nvidia graphics card to run the LGSVL. Please see lecture 1: 1. gitlab.com/ApexAI/autowareclass2020/-/blob/master/lectures/01_DevelopmentEnvironment/devenv.md#prerequisites 2. www.lgsvlsimulator.com/docs/faq/#what-are-the-recommended-system-specs-what-are-the-minimum-required-system-specs
Hi very informative and good lecture about Lidar! Thanks! However I had an error at the end. It seems that my ROS2 Native bridge is not working right? How can I fix this? Could you assist me please? The error is below: I launched the LGSVL simulator with success with a valid vehicle (Lexus1016RXHybrid) and map(San José Parking area). Selected Clusters as Local Machine 127.0.0.1. To set the Simulation I have added the map and vehicle, HOWEVER, it did not work with the suggested bridge by the live class: ROS2 Native, displaying me the following error: Invalid: Incorrect bridge connection string, expected HOSTNAME:PORT Then I selected the "No bridge" and the simulator worked. However when I try to launch the "lidar_bounding_boxes_lgsvl.launch.py", I had the following error: $ ros2 launch autoware_demos /lidar_bounding_boxes_lgsvl.launch.py file '/lidar_bounding_boxes_lgsvl.launch.py' was not found in the share directory of package 'autoware_demos' which is at '/opt/AutowareAuto/share/autoware_demos' Then I have pointed directly to the /opt/AutowareAuto/share folder and authorized the access to the file: **$/opt/AutowareAuto/share/autoware_demos/launch$ sudo chmod 777 lidar_bounding_boxes_**lgsvl.launch.py** ** chmod: changing permissions of 'lidar_bounding_boxes_lgsvl.launch.py': Read-only file system and later launched from there: **$ /opt/AutowareAuto/share/autoware_demos/launch$ ros2 launch /opt/AutowareAuto/share/autoware_demos/launch/ **lidar_bounding_boxes_lgsvl.launch.py malformed launch argument 'lidar_bounding_boxes_lgsvl.launch.py', expected format ':=' So, what is this error? Should I add some name or values besides the launch file? Thanks in advance.
The native_ROS 2 connection I fixed setting the IP address to 127.0.0.1:9090 into Simulations tab interface. However I keep having that launch error. Must I build the pkg? How to do this in ROS2?
@@MarcusVinicius-lq3fe as mentioned in the earlier comment: We built this course hastily due to the sudden emergence of Covid and while we were still in the heavy development of Autoware.Auto. However the team is now days away from the 1.0 release: gitlab.com/autowarefoundation/autoware.auto/AutowareAuto/-/milestones/29 and I would suggest to instead working with the code and documentation on the master branch: autowarefoundation.gitlab.io/autoware.auto/AutowareAuto/usage.html. If you then run into the problems: us the channels mentioned in the autowarefoundation.gitlab.io/autoware.auto/AutowareAuto/support-guidelines.html to get help. If you want to join the Autoware developers, please join our weekly Autonomy Software Working Group meetings: gitlab.com/autowarefoundation/autoware-foundation/-/wikis/Autonomy-Software-working-group.
@@dejanpangercic3367 thanks a lot. I have seen recently in lecture 12 about path planners and autonomous parking. In that time was done in other tool and the guy said it would later integrate within Autoware. Is that already integrated?? I have a lot of interest in test path planners switch them and do a Benchmark. Does Autoware Auto provides me that functionality? Which link could I follow please?
@@MarcusVinicius-lq3fe above mentioned integration has not happened. Everything that Autoware.Auto has regarding planning is in this folder gitlab.com/autowarefoundation/autoware.auto/AutowareAuto/-/tree/master/src/planning and the documentation is here autowarefoundation.gitlab.io/autoware.auto/AutowareAuto/usage.html. More sophisticated planning mechanisms will be ported from here: github.com/tier4/AutowareArchitectureProposal.iv/tree/master/planning.
The SVL simulator does not work anymore. what alternatives do i pursue? please help
Completely reasonable driving by the truck at 01:06:25
Its for testing against overturned trucks (ehem, ehem, Tesla...) lol
Are there any additional method to avoid dust, snow and rain?
I think vertical ray is really reasonable method for object detection, but really sensitive to noise.
Is there a way to use 'autoware' while using 2D lidar sensor? It doesn't have to be HD map, just simple navigation mapping is enough.
It would definitely be possible, however we never really tested something like this. We most of the time work with the 3D sensors like e.g. Velodyne. For your use case I would suggest checking out navigation.ros.org/.
@@dejanpangercic3367 thank you so much :)
How does the ground filtering work with a sloping road? The amount of slope is normally also not a constant.
We have not tested severe slopes on the road. That is the use case in which you would need to use also other sensors (e.g. IMU) and HD map information.
Is Nvidia graphics card compulsory to run lgvsl simulator? I am getting an error ABORTED CORE DUMPED as soon as i ran the simulator. Kindly help
It is compulsory to have Nvidia graphics card to run the LGSVL. Please see lecture 1:
1. gitlab.com/ApexAI/autowareclass2020/-/blob/master/lectures/01_DevelopmentEnvironment/devenv.md#prerequisites
2. www.lgsvlsimulator.com/docs/faq/#what-are-the-recommended-system-specs-what-are-the-minimum-required-system-specs
@@dejanpangercic5134 Thanks for the response. Can i get it in my non nvidia laptop in any way?
@@codebot2235 were you able to run the simulator? I am getting the same error 'Aborted (Core Dumped)' when I try to run the simulator.
Hi very informative and good lecture about Lidar! Thanks!
However I had an error at the end. It seems that my ROS2 Native bridge is not working right? How can I fix this?
Could you assist me please? The error is below:
I launched the LGSVL simulator with success with a valid vehicle (Lexus1016RXHybrid) and map(San José Parking area). Selected Clusters as Local Machine 127.0.0.1. To set the Simulation I have added the map and vehicle, HOWEVER, it did not work with the suggested bridge by the live class: ROS2 Native, displaying me the following error: Invalid: Incorrect bridge connection string, expected HOSTNAME:PORT
Then I selected the "No bridge" and the simulator worked. However when I try to launch the "lidar_bounding_boxes_lgsvl.launch.py", I had the following error: $ ros2 launch autoware_demos /lidar_bounding_boxes_lgsvl.launch.py
file '/lidar_bounding_boxes_lgsvl.launch.py' was not found in the share directory of package 'autoware_demos' which is at '/opt/AutowareAuto/share/autoware_demos'
Then I have pointed directly to the /opt/AutowareAuto/share folder and authorized the access to the file:
**$/opt/AutowareAuto/share/autoware_demos/launch$ sudo chmod 777 lidar_bounding_boxes_**lgsvl.launch.py** ** chmod: changing permissions of 'lidar_bounding_boxes_lgsvl.launch.py': Read-only file system
and later launched from there:
**$ /opt/AutowareAuto/share/autoware_demos/launch$ ros2 launch /opt/AutowareAuto/share/autoware_demos/launch/ **lidar_bounding_boxes_lgsvl.launch.py malformed launch argument 'lidar_bounding_boxes_lgsvl.launch.py', expected format ':='
So, what is this error? Should I add some name or values besides the launch file? Thanks in advance.
The native_ROS 2 connection I fixed setting the IP address to 127.0.0.1:9090 into Simulations tab interface. However I keep having that launch error. Must I build the pkg? How to do this in ROS2?
Finally sovled the issue looking to lecture 11.....I should dowload afrom github some packages and build with colcon build...
@@MarcusVinicius-lq3fe as mentioned in the earlier comment:
We built this course hastily due to the sudden emergence of Covid and while we were still in the heavy development of Autoware.Auto.
However the team is now days away from the 1.0 release: gitlab.com/autowarefoundation/autoware.auto/AutowareAuto/-/milestones/29 and I would suggest to instead working with the code and documentation on the master branch: autowarefoundation.gitlab.io/autoware.auto/AutowareAuto/usage.html.
If you then run into the problems: us the channels mentioned in the autowarefoundation.gitlab.io/autoware.auto/AutowareAuto/support-guidelines.html to get help.
If you want to join the Autoware developers, please join our weekly Autonomy Software Working Group meetings: gitlab.com/autowarefoundation/autoware-foundation/-/wikis/Autonomy-Software-working-group.
@@dejanpangercic3367 thanks a lot. I have seen recently in lecture 12 about path planners and autonomous parking. In that time was done in other tool and the guy said it would later integrate within Autoware. Is that already integrated?? I have a lot of interest in test path planners switch them and do a Benchmark. Does Autoware Auto provides me that functionality? Which link could I follow please?
@@MarcusVinicius-lq3fe above mentioned integration has not happened. Everything that Autoware.Auto has regarding planning is in this folder gitlab.com/autowarefoundation/autoware.auto/AutowareAuto/-/tree/master/src/planning and the documentation is here autowarefoundation.gitlab.io/autoware.auto/AutowareAuto/usage.html.
More sophisticated planning mechanisms will be ported from here: github.com/tier4/AutowareArchitectureProposal.iv/tree/master/planning.
thx for the lesson!!!
Naiceeeee