Jetpack 461 on 128GB , jetbot docker installation failed

[fredeisen1]

Hi,
I have succesfully installed Jetpack 461 and have played a little bit with jetbot-inference(part of AI World)
I like to build a custom jetbot myself, so I use docker to install and failed… JETBOT_BASE_IMAGE not found for 32.7.2 …
Please help how to proceed.

Flashing and running image jetbot_043_nano-4gb-jp45.zip works ok, but no space left to do any experiment.
/dev/mmcblk0p1 displays 24GB, 100% used and my SD is 128GB.
Fixing the size with SDKMANAGER helps, but the installed jetbot installation vanished.
Your help is very appciated to help this newbee
Thanks,
Fred

[AastaLLL]

Hi,

Do you use the Orin platform?
Please noted that you need JetPack 5 to get the Orin support.

Thanks.

[fredeisen1]

Hi,
Thanks for your response. No, I do not use Orin. I donot even know what is.
I use the SDKMANAGER and Etcher to do the software.

Now I have tried to install JETBOT.
1)
First I have flashed Jetpack461(and later Jetpack451) and then use the docker’s jetbot installation. Both failed at ’ cd docker ./enable.sh $HOME’
What is the correct way to get it done without errors?
2)
I have flashed the jetbot image ‘jetbot-043_nano-4gb-jp45.zip’ and it is up and running. However on the 128GB and 32GB I have got hardly space to do any sensible experiments. ‘dh -h’ displayed 24GB (92%full) on both the SD cards.
How can I fix this, sothat the actual size will be used.

Hope to receive guidance to solve the above
Thanks and best regards,
Fred

[AastaLLL]

Hi,

Which Jetson device do you use?
If you don’t mind, we will move your topic to the corresponding board.

Thanks.

[fredeisen1]

Hi,

It is a Jetson Nano Developerkit 4GB

Best regards,

Fred

[AastaLLL]

Hi,

Could you try to reflash the JetPack 4.6.1 to see if the 128G SD card can be recognized first?
Thanks.

[fredeisen1]

Hi,
Jetpack 461 on 128GB is up and running!!!
Also the Jetpack 451.
But as I said I like to play around with the Jetbot.
Using the available docker’s Jetbot failed on both the Jetpacks.
What is your advise here?
Thanks

[naisy]

You can try our custom jetbot image.
https://faboplatform.github.io/JetbotDocs/02.ソフトウェア設定/01.Jetson%20Nano/02.OSアップグレード手順/03.jetbot/

You can run the following functions

• basic_motion
  • 01_basic_motion_JP.ipynb
• collision_avoidance
  • 01_data_collection_JP.ipynb
  • 02_train_model_resnet18_JP.ipynb
  • 03_live_demo_resnet18_build_trt_JP.ipynb
  • 04_live_demo_resnet18_trt_JP.ipynb
• road_following
  • 01_data_collection_JP.ipynb
  • 02_train_model_JP.ipynb
  • 03_live_demo_build_trt_JP.ipynb
  • 04_live_demo_trt_JP.ipynb
• object_following
  • 01_build_ssd_mobilenet_v2_fpnlite_JP.ipynb
  • 02_live_demo_JP.ipynb

Remember to run camera.stop() at the end of the cell before exiting. If you forget to do this, the camera will stop working. In that case, reboot jetson nano.
When closing the notebook, please shut down the Jupyter Kernel to free up memory.

[fredeisen1]

Hi Naisy,

Thank you for uour response.

Several attemps made to download jetbot-nano-4gb-jp461-2022-04-04.img.zip (13GB)

All failed with the message ¡®network error ¨C failed , all nearly at the end.

I do not know if I got the permission to download that file.

I will try again tomorrow.

Regards,

Fred

[fredeisen1]

Hi Naisy,
Finally I got my 128GB flashed the jetbot-nano-4gb-jp461-2022-04-04.img.zip image.
Jetbot is operational now.
Thanks for your help
Cheers,
Fred

[naisy]

I recently had a request from one of our clients to combine object following and road following, so I have uploaded the code.
The jetbot running according to road following will stop when it detects the target. (default: person)
You can use this as well.

Put ipynb file in the object detection directory.

wget https://raw.githubusercontent.com/naisy/jetbot/master/notebooks/object_following/02_live_demo_road_following_JP.ipynb

Have fun!

[fredeisen1]

Hi Naisy,

Great! Thank you very much for recommendation and your help

Cheers,

Fred

[fredeisen1]

Hi Naisy,
The jetbot-nano-4gb-jp461-2022-04-04.img.zip image I got from you doesnot have the directory object detection where I can download.
There is only a docker directory. Others are empty.

Thus I can not use the suggestion below:
wget https://raw.githubusercontent.com/naisy/jetbot/master/notebooks/object_following/02_live_demo_road_following_JP.ipynb

Can I get and download the corresponding jetbot folders somewhere?
Regards, Fred

[naisy]

1. Wi-Fi setup.

JetBot uses JupyterLab. Operation is done through a browser.
Therefore, Wi-Fi setup is required first.

sudo nmcli device wifi connect 'MySSID' password 'MyPassword' ifname wlan0

# check your jetson ip address
sudo ifconfig -a 

https://faboplatform.github.io/JetbotDocs/02.ソフトウェア設定/01.Jetson%20Nano/01.設定手順/02.Wi-Fi設定/01.network_hdmi_nogui/

---

2. Access jupyterlab in your pc browser.

Then access JupyterLab in your browser.
IP address should be your JetBot’s IP address.
The JupyterLab port number is 8888.
The password is jetbot.

http://192.168.xxx.xxx:8888

https://faboplatform.github.io/JetbotDocs/05.Basic%20Motion/01.run/

---

3. You can run jupyter notebooks.

You will find the .ipynb file in the following directories.

• basic_motion
• collision_avoidance
• road_following
• object_following

---

Other information.

This JetBot is running on Docker.
The directory path on browser is /Notebooks in Jupyterlab.
The directory path in the docker container is ~/Notebooks.

You can launch a virtual terminal in Jupyterlab.

cd ~/Notebooks/object_following
wget https://raw.githubusercontent.com/naisy/jetbot/master/notebooks/object_following/02_live_demo_road_following_JP.ipynb

or, download that file to you pc, and upload in the object_following directory (use jupyterlab upload button).

[fredeisen1]

Hi Naisy,
Thanks again for your guidance.
My US-keyboard is not supported by this jetbot image.
I can not find the Asterisks and Small Tilde.
Maybe you can help me to correct it.
Regards, Fred

[naisy]

It is the same as general ubuntu.
https://wiki.debian.org/Keyboard#Manual_configuration_of_keyboard

Check current keyboard settings

cat /etc/default/keyboard

Edit keyboard settings

sudo vi /etc/default/keyboard

before:

# KEYBOARD CONFIGURATION FILE

# Consult the keyboard(5) manual page.

XKBMODEL="pc105"
XKBLAYOUT="jp"
XKBVARIANT="OADG109A"
XKBOPTIONS=""

BACKSPACE="guess"

after:

# KEYBOARD CONFIGURATION FILE

# Consult the keyboard(5) manual page.

XKBMODEL="pc105"
XKBLAYOUT="us"
XKBVARIANT=""
XKBOPTIONS=""

BACKSPACE="guess"

Apply keyboard settings

sudo dpkg-reconfigure keyboard-configuration

[fredeisen1]

Hi Naisy,
Thanks for your help and your patience.
Cheers, Fred