yes i agree. i wanted to try that out as well but didn't find the time yet. i actually received an e-mail from nvidia a few days ago where they talk about exactly this: developer.nvidia.com/blog/accelerating-ai-modules-for-ros-and-ros-2-on-jetson/?ncid=em-news-556766&mkt_tok=MTU2LU9GTi03NDIAAAF9LtsHY5qwpxFPMWgfekTrYhSo8wGYWOvTP82GK6nc6BBJvl-oUApAvowx8xiastda8i8enM4qke9FIqVtF0lUHRJ-OtRvS-ka_q4DC-5YlcTIch-X#cid=em07_em-news_en-us
Hmm I didn't face this issue. Maybe try if you can get into a console with Ctrl+Alt+F3 and see if you can find any information like that. Other problems that I can think of would be a corrupt image (try flashing it again) or an invalid configuration (e.g. a typo in the extlinux.conf).
Great work, can you please explain how to install cuda, TF and opencv for object detection on this ubuntu 20.04 ? How much RAM optimization is there with this?
Hi sir, I'm using A02 version. When I flash the sd card and insert it in to jetson it doesn't happen anything. It's just a blank screen. Can you show me some solution?
Hello When I do that procces jetpack coming with together ? or all will be brake ? is there any solution ? using jetpack with not ubuntu 18 only with 20.04?
Firstly, thank you so much! Have been using your image without any issue. But I need some help. I am unable to install tensorflow. It’s no matching distribution found. Can it be a python version issue? Any suggestion will be appreciated.
I haven't tried installing Tensorflow. In general I'd follow this guide: docs.nvidia.com/deeplearning/frameworks/install-tf-jetson-platform/index.html They write that Jetpack is required, I'm not sure if you can install it on the image, you'll just have to try (docs.nvidia.com/jetson/jetpack/install-jetpack/index.html#package-management-tool). If you can't make it work with the image there could also be the possibility to run a docker container with ROS2 on it and use one of the official Nvidia images for the Jetson Nano (developer.nvidia.com/blog/accelerating-ai-modules-for-ros-and-ros-2-on-jetson/?ncid=em-news-556766&mkt_tok=MTU2LU9GTi03NDIAAAF9LtsHY5qwpxFPMWgfekTrYhSo8wGYWOvTP82GK6nc6BBJvl-oUApAvowx8xiastda8i8enM4qke9FIqVtF0lUHRJ-OtRvS-ka_q4DC-5YlcTIch-X#cid=em07_em-news_en-us).
@@roboticsinanutshell thank you for your response. I followed Nvidia’s tf installing guide but it didn’t work. It must be due to jetpack, I tried installing jetpack but that didn’t work. I might try docker but docker has a learning curve and i’m short on time. I need ROS noetic. I found a ROS noetic build from source tutorial for ubuntu 18, I might try that. Thank you again.
Wanted to mention that, after booting it, things go well. However, when running apt update and apt upgrade, things get destroyed. Do you happen to have other .iso versions?
The sudo apt update && sudo apt install locales process takes way to long when I tried to install Ros2 foxy, don't know what is going on - it says: Waiting for cash lock: Could not get lock /var/lib/dpkg/lock-frontend. Is held by process 5851 (unattended-upgr) - have you gotten this issue?
@@ruanpotgieter2596 To be honest, after performing the upgrade, the entire apt was destroyed. I narrowed it down to one package: linux-firmware. It could be, from what I read, my hardware is already not compatible, then the upgrade is kind of left in the air not knowing what to do. To your question, I do get the same error, but please let's not assume is the same cause. Please determine first, that there is no other process running in parallel the apt update or apt install. Check each package when the errors are thrown, etc. Try to determine the cause of your errors by reading the logs too.
@@walterp773 good day, thanks for responding, I have stopped the process running in parallel and it seemed to work, how can I check the logs on the jetson nano?
For my purposes I used the Jetson Nano B01 with 4 GB RAM. Other users in the comments have tried the 2GB version but I think no one confirmed yet that it works (one described that he can see a "strange screen" though which indicates that he might have been able to boot or at least get into the boot sequence). So it can not be guaranteed that it's possible on the 2GB version (yet). Maybe you have to do the step that is described at 2:12 but have to use a different FDT entry (see 3:14). There is some info in the video on how to find out which FTD entry you have to use. You should also make sure the flashing worked as expected (see conversation in the comments with Daniel Couser). There are also other possible approaches. NVIDIA has released a docker image that could solve the problem, I didn't try that though (developer.nvidia.com/blog/accelerating-ai-modules-for-ros-and-ros-2-on-jetson/?ncid=em-news-556766&mkt_tok=MTU2LU9GTi03NDIAAAF9LtsHY5qwpxFPMWgfekTrYhSo8wGYWOvTP82GK6nc6BBJvl-oUApAvowx8xiastda8i8enM4qke9FIqVtF0lUHRJ-OtRvS-ka_q4DC-5YlcTIch-X#cid=em07_em-news_en-us ). Another user in the comments said that he upgraded the installation from the latest and original Nvidia image all the way to 20.10. while maintaining functionality. So there are different options. You could be the first to confirm one of these approaches to work. So I wish you best of luck! Let us know if you can make it work. :) Cheers Fynn
What did you use to extract the file? I tried using Winrar. It would show the flash image but give me errors when I went to get the file. (do you know of any extractor apps that can be downloaded on windows or a link to an already unzipped file?
Hi there, it seems that Windows 10 might support unzipping these files directly from the console (pureinfotech.com/extract-tar-gz-files-windows-10/). Otherwise you could also give 7-Zip a chance. To give you a lift I uploaded the unzipped file to a google drive where you can download it: drive.google.com/file/d/1O9uzztQPAJit1gLDkE-5O6biRt7dY5yG/view?usp=sharing Let me know once you downloaded it, then I will free up the space again.
7Zip will work if your on windows. You may need to extract twice, the first time you'll get a .tar.tar file, then the second time you'll get the disk image file
If you want to use ROS 2 I don't think it will suffice if you have Ubuntu 16. As far as I know you need at least Ubuntu 18 and build from source (e.g. with these Docker containers: developer.nvidia.com/blog/accelerating-ai-modules-for-ros-and-ros-2-on-jetson/?ncid=em-news-556766&mkt_tok=MTU2LU9GTi03NDIAAAF9LtsHY5qwpxFPMWgfekTrYhSo8wGYWOvTP82GK6nc6BBJvl-oUApAvowx8xiastda8i8enM4qke9FIqVtF0lUHRJ-OtRvS-ka_q4DC-5YlcTIch-X#cid=em07_em-news_en-us) or you need Ubuntu 20 and can install prebuilt binaries of ROS 2 (sudo apt install).
This was tested on a Jetson Nano with 4GB RAM. I haven't tested it on the new 2GB Nano. I can imagine that this will also work on the 2GB version but I can't say for sure.
@@roboticsinanutshell actually my nano was broke while it got delivered then it took 2 weeks to get it replaced, got it few days back. Will try it in few days , btw nano 2gb image is based on Ubuntu 16
@@eminemhc5763 good to hear you were able to get it replaced. let me know how it goes if you try it. :) when you say that the 2gb image is based on ubuntu 16, are you referring to the official nvidia image?
I tried to flash the img onto my sd card, but afterwards I can not find the sd card anymore. I need to edit the one file because I have the model B01. Do you know what might be happening when I flash the file? Any thoughts on how to fix this? This tutorial is exactly what I have been looking for, thank you!
Hi there, hmm that problem didn't occur to me. Did you use Balena Etcher to flash it or something else? Is your SD card sufficiently big? I used one with 32 GB. Maybe the device you use for flashing doesn't support mounting the SD card? Steps that come to my mind: - Unplug SD card stick and replug it - Format SD card before flashing and try again - Try another SD card - Try another device to flash / mount the SD card
Hii , i tried the same steps for installation. The jetson nano is not booting up . The led on jetson is green yet the display only shows nvidia and loops again. i tried flashing other ubuntu 18 (developer.nvidia.com/embedded/learn/get-started-jetson-nano-devkit#write)image into sd card it is able to boot.
hi there, this problem was reported by multiple people now. This problem did not occur to me so it's difficult to say what the issue is. I used the Jetson Nano B01 with 4 GB RAM. Are you using the 2 GB version or the A02 Jetson?
@@roboticsinanutshell I scrapped the attempt to get it to work. It is easier just to upgrade the installation from the latest and original Nvidia image. I was actually able to upgrade all the way to 20.10 and still maintain functionality.
![Noob To Max With DRAGON REWORK In Blox Fruits [FULL MOVIE]](http://i.ytimg.com/vi/LnBMOinoOvA/mqdefault.jpg)
![Felix "Unfair" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/Oswujxm2Ag0/mqdefault.jpg)
Worked flawlessly :-)
I haven't tried it myself, but I think you can easily use docker with the appropriate images to archive the same result.
yes i agree. i wanted to try that out as well but didn't find the time yet. i actually received an e-mail from nvidia a few days ago where they talk about exactly this: developer.nvidia.com/blog/accelerating-ai-modules-for-ros-and-ros-2-on-jetson/?ncid=em-news-556766&mkt_tok=MTU2LU9GTi03NDIAAAF9LtsHY5qwpxFPMWgfekTrYhSo8wGYWOvTP82GK6nc6BBJvl-oUApAvowx8xiastda8i8enM4qke9FIqVtF0lUHRJ-OtRvS-ka_q4DC-5YlcTIch-X#cid=em07_em-news_en-us
Hello Sir,
I did all the step and the start my Jetson Nano LT4 B01 but once it starts it just showing me the logo of Nvidia
Hmm I didn't face this issue. Maybe try if you can get into a console with Ctrl+Alt+F3 and see if you can find any information like that.
Other problems that I can think of would be a corrupt image (try flashing it again) or an invalid configuration (e.g. a typo in the extlinux.conf).
Great work, can you please explain how to install cuda, TF and opencv for object detection on this ubuntu 20.04 ? How much RAM optimization is there with this?
Hi sir, I'm using A02 version. When I flash the sd card and insert it in to jetson it doesn't happen anything. It's just a blank screen. Can you show me some solution?
Good day, My xubuntu image works on my jetson B01 dev kit but the firefox browser is causing random glitches on my monitor. How do I fix this?
Hello When I do that procces jetpack coming with together ? or all will be brake ? is there any solution ? using jetpack with not ubuntu 18 only with 20.04?
I want to do this, but would I still be able to install CUDA, tensorrt, and things of that nature that come with Jetpack?
Did you manage a method to do this?
Firstly, thank you so much! Have been using your image without any issue. But I need some help. I am unable to install tensorflow. It’s no matching distribution found. Can it be a python version issue? Any suggestion will be appreciated.
I haven't tried installing Tensorflow. In general I'd follow this guide: docs.nvidia.com/deeplearning/frameworks/install-tf-jetson-platform/index.html
They write that Jetpack is required, I'm not sure if you can install it on the image, you'll just have to try (docs.nvidia.com/jetson/jetpack/install-jetpack/index.html#package-management-tool). If you can't make it work with the image there could also be the possibility to run a docker container with ROS2 on it and use one of the official Nvidia images for the Jetson Nano (developer.nvidia.com/blog/accelerating-ai-modules-for-ros-and-ros-2-on-jetson/?ncid=em-news-556766&mkt_tok=MTU2LU9GTi03NDIAAAF9LtsHY5qwpxFPMWgfekTrYhSo8wGYWOvTP82GK6nc6BBJvl-oUApAvowx8xiastda8i8enM4qke9FIqVtF0lUHRJ-OtRvS-ka_q4DC-5YlcTIch-X#cid=em07_em-news_en-us).
@@roboticsinanutshell thank you for your response. I followed Nvidia’s tf installing guide but it didn’t work. It must be due to jetpack, I tried installing jetpack but that didn’t work. I might try docker but docker has a learning curve and i’m short on time. I need ROS noetic. I found a ROS noetic build from source tutorial for ubuntu 18, I might try that. Thank you again.
Wanted to mention that, after booting it, things go well.
However, when running apt update and apt upgrade, things get destroyed.
Do you happen to have other .iso versions?
The sudo apt update && sudo apt install locales process takes way to long when I tried to install Ros2 foxy, don't know what is going on - it says: Waiting for cash lock: Could not get lock /var/lib/dpkg/lock-frontend. Is held by process 5851 (unattended-upgr) - have you gotten this issue?
@@ruanpotgieter2596 To be honest, after performing the upgrade, the entire apt was destroyed. I narrowed it down to one package: linux-firmware. It could be, from what I read, my hardware is already not compatible, then the upgrade is kind of left in the air not knowing what to do.
To your question, I do get the same error, but please let's not assume is the same cause. Please determine first, that there is no other process running in parallel the apt update or apt install. Check each package when the errors are thrown, etc. Try to determine the cause of your errors by reading the logs too.
@@walterp773 good day, thanks for responding, I have stopped the process running in parallel and it seemed to work, how can I check the logs on the jetson nano?
@@ruanpotgieter2596 no worries.
Logs for apt? Think you can run:
cat /var/log/apt/history.log
i have tried install it on my nano 2GB, it didnot work(it cannot even boot up), any suggestions?
For my purposes I used the Jetson Nano B01 with 4 GB RAM. Other users in the comments have tried the 2GB version but I think no one confirmed yet that it works (one described that he can see a "strange screen" though which indicates that he might have been able to boot or at least get into the boot sequence). So it can not be guaranteed that it's possible on the 2GB version (yet).
Maybe you have to do the step that is described at 2:12 but have to use a different FDT entry (see 3:14). There is some info in the video on how to find out which FTD entry you have to use.
You should also make sure the flashing worked as expected (see conversation in the comments with Daniel Couser).
There are also other possible approaches. NVIDIA has released a docker image that could solve the problem, I didn't try that though (developer.nvidia.com/blog/accelerating-ai-modules-for-ros-and-ros-2-on-jetson/?ncid=em-news-556766&mkt_tok=MTU2LU9GTi03NDIAAAF9LtsHY5qwpxFPMWgfekTrYhSo8wGYWOvTP82GK6nc6BBJvl-oUApAvowx8xiastda8i8enM4qke9FIqVtF0lUHRJ-OtRvS-ka_q4DC-5YlcTIch-X#cid=em07_em-news_en-us ).
Another user in the comments said that he upgraded the installation from the latest and original Nvidia image all the way to 20.10. while maintaining functionality.
So there are different options. You could be the first to confirm one of these approaches to work. So I wish you best of luck! Let us know if you can make it work. :)
Cheers
Fynn
Hi, have you tried any of the solutions?
What did you use to extract the file? I tried using Winrar. It would show the flash image but give me errors when I went to get the file. (do you know of any extractor apps that can be downloaded on windows or a link to an already unzipped file?
Hi there, it seems that Windows 10 might support unzipping these files directly from the console (pureinfotech.com/extract-tar-gz-files-windows-10/). Otherwise you could also give 7-Zip a chance. To give you a lift I uploaded the unzipped file to a google drive where you can download it: drive.google.com/file/d/1O9uzztQPAJit1gLDkE-5O6biRt7dY5yG/view?usp=sharing
Let me know once you downloaded it, then I will free up the space again.
I extracetd the .iso using 7zip and wrote that image to the sdcard using the RaspberryPi Imager sw.
7Zip will work if your on windows. You may need to extract twice, the first time you'll get a .tar.tar file, then the second time you'll get the disk image file
can someone help me my wifi is not working after the boot Intel Corporation Wireless 7265 (rev ff)
Can I use ROS2 Humble on this ubuntu?
Hey, can i flash image onto usb and create a bootable pen drive to install OS on sd card
I haven't done that yet, so I don't know :l
Only we can on ubuntu 20?
Is it ok if it’s ubuntu 16?
If you want to use ROS 2 I don't think it will suffice if you have Ubuntu 16. As far as I know you need at least Ubuntu 18 and build from source (e.g. with these Docker containers: developer.nvidia.com/blog/accelerating-ai-modules-for-ros-and-ros-2-on-jetson/?ncid=em-news-556766&mkt_tok=MTU2LU9GTi03NDIAAAF9LtsHY5qwpxFPMWgfekTrYhSo8wGYWOvTP82GK6nc6BBJvl-oUApAvowx8xiastda8i8enM4qke9FIqVtF0lUHRJ-OtRvS-ka_q4DC-5YlcTIch-X#cid=em07_em-news_en-us) or you need Ubuntu 20 and can install prebuilt binaries of ROS 2 (sudo apt install).
Is this for the new nano 2Gb ???
This was tested on a Jetson Nano with 4GB RAM. I haven't tested it on the new 2GB Nano.
I can imagine that this will also work on the 2GB version but I can't say for sure.
@@roboticsinanutshell thank you for the reply and tutorial , will try it in the new 2gb
@@eminemhc5763 Hey I was wondering if you actually tried it on the 2Gb Jetson Nano. If yes, did it work?
@@roboticsinanutshell actually my nano was broke while it got delivered then it took 2 weeks to get it replaced, got it few days back. Will try it in few days , btw nano 2gb image is based on Ubuntu 16
@@eminemhc5763 good to hear you were able to get it replaced. let me know how it goes if you try it. :) when you say that the 2gb image is based on ubuntu 16, are you referring to the official nvidia image?
I tried to flash the img onto my sd card, but afterwards I can not find the sd card anymore. I need to edit the one file because I have the model B01. Do you know what might be happening when I flash the file? Any thoughts on how to fix this?
This tutorial is exactly what I have been looking for, thank you!
Hi there,
hmm that problem didn't occur to me. Did you use Balena Etcher to flash it or something else?
Is your SD card sufficiently big? I used one with 32 GB.
Maybe the device you use for flashing doesn't support mounting the SD card?
Steps that come to my mind:
- Unplug SD card stick and replug it
- Format SD card before flashing and try again
- Try another SD card
- Try another device to flash / mount the SD card
did you flash using windows or Linux? Mine boots ok, but I can't edit the ext partitions on Windows without using something like ext2fs
Did you fix this problem somehow? Im having the same issue... :(
It only reads it on Linux
Thank you ... super interesting.... can we keep in touch by email? i have to ask you for a collab...
Hi there, you can reach us at roboticsinanutshell@gmail.com
Hii , i tried the same steps for installation. The jetson nano is not booting up . The led on jetson is green yet the display only shows nvidia and loops again. i tried flashing other ubuntu 18 (developer.nvidia.com/embedded/learn/get-started-jetson-nano-devkit#write)image into sd card it is able to boot.
hi there, this problem was reported by multiple people now.
This problem did not occur to me so it's difficult to say what the issue is.
I used the Jetson Nano B01 with 4 GB RAM. Are you using the 2 GB version or the A02 Jetson?
Same issue here as well. After initial Ubuntu Configuration, it doesn't boot afterwards.
@@zacharygoldsmith7218 did you use the Jetson Nano B01 with 4GB or something else?
@@roboticsinanutshell I scrapped the attempt to get it to work. It is easier just to upgrade the installation from the latest and original Nvidia image. I was actually able to upgrade all the way to 20.10 and still maintain functionality.
@@roboticsinanutshell And yes. It's a Jetson Nano B01 4GB