This guide provides a comprehensive, step-by-step instruction on setting up your Jetson Nano with Ubuntu 20.04. The guide covers the preparation, downloading and flashing of the image, and setting up Jetson Nano, along with troubleshooting tips, ROS2 installation, and additional setup scripts.
Before you begin, make sure you have the following items:
- Jetson Nano board
- Minimum 32 GB microSD card (recommended: 64 GB or larger for more space)
- Card reader (for flashing the image to the SD card)
- Computer (running Windows, macOS, or Linux)
- Jetson Nano power supply (5V 4A recommended)
- Keyboard, mouse, HDMI monitor (for setup)
Visit the link to download the Jetson Nano Ubuntu 20.04 image with pre-installed OpenCV, TensorFlow, and PyTorch. The file name is JetsonNanoUb20_3b.img.xz and it is 8.7 GB in size. Use either Sync or Google Drive options. Google Drive has daily download limits, so use it only if Sync is slow.
- Get a 32 GB or larger SD card. A 64 GB or larger card is recommended for extra working space.
To install Balena Etcher for flashing SD card images, run the following script:
./install-balena-etcher.sh- Use a tool like Balena Etcher to flash the
.img.xzfile directly onto the SD card. - Note: Do not unzip the
.img.xzfile. Flash it as is. - md5sum: D738F1FE20088A1BDBD10E2358B512F7
After booting, your SD card's storage may be nearly full because of all the pre-installed software. This step is optional but recommended.
Install GParted
- Open a terminal and run:
sudo apt-get install gpartedResize the Partition:
- Launch GParted and use it to extend the main partition to the full size of your SD card (recommended if you used a 64 GB or larger card).
This image comes pre-installed with the following tools:
- OpenCV 4.8.0: For computer vision tasks
- TensorFlow 2.4.1: For deep learning
- PyTorch 1.13.0 and TorchVision 0.14.0: For neural network applications
- TensorRT 8.0.1.6: Optimizing deep learning inference
- TeamViewer and Jtop: For remote control and system monitoring
When connecting to the internet for the first time, allow Software Updater to complete updates for your system to ensure all components are up-to-date.
The Chromium browser can interfere with some system processes. Instead, use the pre-installed Firefox browser.
The SD card comes filled with software (around 21 GB). Ensure you use at least a 64 GB card for additional storage space during your projects.
You might see a Vulkan-related error message during boot. To resolve it, remove the Vulkan folder:
sudo rm -rf /usr/share/vulkan/icd.dIf importing OpenCV and TensorFlow throws an error related to memory allocation, import OpenCV first in your Python scripts.
If an upgrade fails due to a conflicting sleep.conf file, follow the instructions on the linked webpage for resolution.
Below are some helpful scripts to automate the installation of key components for your Jetson Nano setup:
To install Docker Engine for containerized application management, run:
./install-docker-engine.shFor more details on Docker installation, visit the official Install Docker Engine documentation.
For setting up ROS2 on Jetson Nano, run:
./install-ros2.shThis will guide you through the process of installing the latest ROS2 version compatible with Ubuntu 20.04.
If you want to work with ROS2, this image is compatible with it. A comprehensive guide on setting up Jetson Nano with ROS2:
Jetson Nano can be generally powered from a micro USB power supply (5V 2A 10W max). This can manage a keyboard, a mouse, and a small camera. If the planned use case for Jetson Nano is to run Neural Networks along with Depth cameras, it is better to power the device via the DC barrel jack (5V 4A 20W max).
The system setup comparison section contains various combinations of Ubuntu and ROS2 that were tested with and without GUI, and with and without Docker. The main purpose was to identify the stable combination that allows for the latest ROS version to be used as well as get the maximum output from the Jetson Nano.