I have been trying to configure the RealSense Cameras on a Unitree B1. After connecting to the web and running sudo apt-get update and sudo apt-get upgrade, I tried to install the required packages following the RealSense guide for Jetson NX. RealSense.
I then got the following error related to the jetpack install. Jetpack error.
How do I resolve this?
Dear @abuynits,
The error you have encountered is due to upgrading the Nvidia jetson. As a general rule you shouldn’t upgrade their core kernels or firmware unless following a manual or are experienced with the Nvidia micro computers.
I tried to install the required packages following the RealSense guide for Jetson NX. RealSense.
The binary installation of the realsense is correct and should work without problem.
I then got the following error related to the jetpack install. Jetpack error. How do I resolve this?
Try getting the same Nvidia Os e.g. L4T 32.4 and launch repair. Usually, this issue is very difficult to resolve, and would normally require reinstalling the OS as it is faster.
To resolve this would actually be through the Nvidia forums as this is a jetson exclusive error due to the upgrade command. Technically, all other items should function normally, and for the most part the system should continue to run without a problem.
We usually recommend leaving the Unitree robot’s native computers as is and installing your own computer on top as the Unitree robots, have a lot of scripts running in the background, and a fresh installation would disrupt the system.
Thanks for the response! Is there a way of installing an older L4T version without loosing all of my user-installed files? Unitree installed custom packages that control the lidar and realsense cameras which I don’t want to lose.
Thanks again for the help!
Dear @abuynits,
The process is complicated. I would initially advise keeping the system as it is, if it functions without issues. Altering and repairing the Nvidia usually creates more issues than solving them.
Otherwise, the alternative is (Not Recommended) is to create a backup of the Nvidia board and then note the L4T versions and repair the base kernel image.
Note: Backing up and repairing if done incorrectly can corrupt your card and you may lose all the information in it. Please do not perform these actions if you are not familiar with the procedures.