Question about how to use arena-rosnav with AIO #45
Comments
|
You are right, arena-rosnav does not include aio package atm, but the way you downloaded it is completely fine. Could you let me know the exact command you used to launch aio ? Also, you should be seeing some error messages in your terminal while launching, please send a screenshot as well if you notice any of them. |
|
Hi @Jacenty00, I recently tried to install it on new Docker container, but I failed. So I will show the result after launching start_arena launch file and command. 1. start_arena.launch with aio
I am wondering if I am doing right following the instruction. |
|
Detailed shell codes are here. |
|
After I found that Cuda has problem, I re-setup on the new Docker container using related cuda version. Environment is : But, then, I couldn't catkin_make. 1. 2. 3. |
Yeah, everything looks fine here. These are actually warning messages and can be ignored most of the time. If you want to get rid of them, you would have to lower the values of controller_frequency and/or costmap update frequency. These can be found here: |
|
Regarding the actual problem, having the Docker image cuda version match your driver is the right choice. I just did the installation process using the same docker image you referenced and was able to get AiO running without any errors. My guess would be that you are missing some installation step. Make sure you follow our install guide. Probably some required ROS packages are missing, could be the reason why you are not able to run catkin_make. If you have problems while installing our fork of stable-baselines3, making sure that you are using specific version of setuptools may be useful pip install --upgrade setuptools==65.5.0 |
Also, whenever catkin_make fails to compile, it should provide some more useful error message as to which package was the culprit. It may be that you need to scroll through the terminal, in order to find it. |
|
Thank you for your help and answer.
After following the guide, I had the same error below. Thank you for your help sincerely. ========================================== |
|
It looks like the error is coming from one of the packages in /root/catkin_ws/utils/3rd-party-robot-packages. You could try deleting that folder and see whether that resolves the issue. But it seems to me like the real cause is something else, as I had never encountered this error before. How did you install ROS inside docker ? Did you use the noetic-desktop-full version (That is what I've used so far) ? |
What exactly did you change here? Your Cuda version or your PyTorch version? I would highly recommend you to not downgrade cuda but instead install the correct pytorch version. |
Yes I made a container using Cuda12.0 Docker image and installes noetic-desktop-full version. |
Hi, I only changed Docker’s cuda version from cudagl-11.04 to cuda12.0 to match with my PC’s. |
|
@doyeon-danny |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Hi @ReykCS .
I want to use All-In-One DRL planner in Arena-rosnav environment, but I got some problems.
I think Arena-rosnav doesn't include AIO package, so I found different guide and added AIO package in .rosinstall file.
And it didn't have arena-bench as well.
After installing AIO package, I launched start_arena.launch file with AIO local planner, but it didn't work.
Just launching default start_arena.launch worked.
I am trying it in my docker environment with Ubuntu 20.04 and Noetic ROS version.
I want to set up your environment to test and compare new planning algorthms.
Can you give me some tips how to set up your environment?
Thank you.
The text was updated successfully, but these errors were encountered: