Aerial Robotics IITK
  • Introduction
  • Danger Zone
  • Tutorials
    • Workspace Setup
      • Installing Ubuntu
      • Basic Linux Setup
      • Spruce up your space
      • ROS Setup
      • PX4 Setup
        • PX4 Toolchain Setup
      • Ardupilot Setup
      • Installing Ground Control Station
        • QGroundControl
        • Mission Planner
      • ArduPilot Setup on Docker
      • PX4 Setup on Docker
    • How to Write a ROS Package
      • ROS Package
      • Node Handles, Parameters, and Topics
      • Coding Standards
      • Custom mavros message
      • Transformations
      • Conversions
    • Cheatsheets
      • CMakeCheatsheet
      • GitCheatsheet
      • LatexCheatsheet
      • Markdown Cheatsheet
    • Miscellaneous
      • Odroid XU4 Setup
      • Simulation using Offboard Control
        • Enable Offboard Mode in PX4
      • Writing a UDev rule
      • Sensor fusion
    • Reference wiki links
  • Concepts
    • Quaternions
      • Theory
    • Kalman Filters
    • Rotations
    • Path Planning
      • Grassfire Algorithm
      • Dijkstra Algorithm
      • A* Algorithm
      • Probabilistic Roadmap
      • RRT Algorithm
      • Visibility Graph Analysis
    • Lectures
      • Aerial Robotics
      • Avionics
      • Control Systems: Introduction
      • Control Systems: Models
      • Inter IIT Tech Meet 2018
      • Kalman Filters
      • Linux and Git
      • Git Tutorial
      • ROS
      • Rotorcraft
      • Software Training
  • Control System
    • Model Predictive Control
      • System Identification
      • Sample SysId Launch Files
      • Running MPC
        • MPC with Rotors
        • MPC with PX4 Sim
        • MPC with ROS
      • References
    • PID Controller
      • Introduction
      • Basic Theory
  • Estimation
    • Visual-Inertial Odometry
      • Hardware Requirements
      • Visual-Inertial Sensing
      • DIYing a VI-Sensor
    • Setup with VICON
    • Odometry from pose data
  • Computer Vision
    • Intel RealSense D435i setup for ROS Noetic
    • IntelRealSense D435i Calibration
    • Camera Calibration
    • ArUco ROS
  • Machine Learning
    • Datasets
  • Hardware Integration
    • Configuring Radio Telemetry
    • Setting up RTK + GPS
    • Integration of Sensors with PixHawk
      • Connecting Lidar-lite through I2C
    • Connections
    • Setting up Offboard Mission
      • Setting up Companion Computer
        • Raspberry Pi 4B Setup
        • Jetson TX2 Setup
      • Communication Setup
      • Guided mode
    • Miscellaneous
  • Resources
    • Open-source algorithms and resources
    • Courses
      • State Space Modelling of a Multirotor
      • Path Planning Lecture
      • Introduction to AI in Robotics
      • RRT, RRT* and RRT*- Path Planning Algorithms
    • Useful Reading Links
      • Aerial Robotics
      • Books
      • Computer Vision and Image Processing
      • Courses on AI and Robotics
      • Deep Neural Network
      • Dynamics and Controls system
      • Motion Planning
      • Probabilistic Robotics
      • Programming
      • Robotics Hardware
      • Miscellaneous and Awesome
    • Online Purchase websites
  • Competitions
    • Inter-IIT TechMeet 8.0
    • Inter-IIT TechMeet 9.0
    • IMAV 2019, Madrid, Spain
    • Inter-IIT TechMeet 10.0
    • Inter-IIT TechMeet 11.0
Powered by GitBook
On this page

Was this helpful?

  1. Hardware Integration
  2. Setting up Offboard Mission
  3. Setting up Companion Computer

Jetson TX2 Setup

PreviousRaspberry Pi 4B SetupNextCommunication Setup

Last updated 1 year ago

Was this helpful?

This guide demonstrates how to flash Jetson TX2 using Docker on Ubuntu 22.04. Docker provides a convenient environment for executing the NVIDIA SDK Manager tool without worrying about dependencies or system configurations.

Prerequisites

  • Docker installed on the host PC

  • Jetson TX2 development kit

  • Internet connection

Steps

  1. Download Docker Image

    • Download the Docker image from to your host PC.

  2. Load Docker Image

    • Navigate to the directory where the downloaded Docker image is located.

    • Run the following command to load the Docker image:

      docker load -i ./sdkmanager-[version].[build#]-[base_OS]_docker.tar.gz

      Replace [version].[build#] and [base_OS] with the appropriate values. Press tab for autofill.

  3. Tag the Version (Optional)

    • It's recommended to tag the version as "latest" for ease of use:

      docker tag sdkmanager:[version].[build#] sdkmanager:latest
  4. Flash Jetson TX2

    • Run the following command to initiate the flashing process

      docker run -it --privileged -v /dev/bus/usb:/dev/bus/usb/ -v /dev:/dev -v /media/$USER:/media/nvidia:slave --name JetPack_TX2_Devkit --network host sdkmanager --cli --action install --login-type devzone --product Jetson --target-os Linux --version 4.6.4 --host --target JETSON_TX2_TARGETS --flash --license accept --stay-logged-in true --collect-usage-data enable --exit-on-finish

      This command initiates the SDK Manager in CLI mode and starts the flashing process.

    • Once the installation and flashing are complete, commit the container as a new image:

      docker commit JetPack_TX2_Devkit jetpack_tx2_devkit:flash
    • Remove the original container:

      docker container rm JetPack_TX2_Devkit
    • Run the flash operation directly from the created image:

      docker run -it --rm --privileged -v /dev/bus/usb:/dev/bus/usb/ jetpack_tx2_devkit:flash

    PS: You can use the 1st command again without the --name tag

  5. Follow On-screen Instructions

    • Choose the packages to be installed.

    • Ignore any low memory warnings and proceed.

    • Select either automatic or manual mode.

    • Choose automatic mode, follow the on-screen instructions to complete the flashing process.

  6. Completion

    • Once the flashing process is completed, your Jetson TX2 will be flashed and ready for use.

Additional Notes

  • Make sure to connect the Jetson TX2 development kit to your host PC via USB during the flashing process.

With this Docker-based approach, flashing your Jetson TX2 becomes a seamless process, eliminating the hassle of managing dependencies and configurations.


Known Issues and Workarounds for SDK Manager Docker Image

Issue:

When installing JetPack SDK using the SDK Manager Docker image, you may encounter the following error during installation from the File System and OS component:

dpkg: Exec format error

Workaround:

Before installing JetPack SDK, ensure that qemu-user-static and binfmt-support are installed on your host machine. Run the following command on your host machine:

sudo apt-get install qemu-user-static binfmt-support
sudo update-binfmts --enable

DRIVE SDK Installation

Issue:

Flashing the target may fail during DRIVE SDK installation if ADB is running on the host machine.

Workaround:

Before flashing the target, make sure to stop ADB using the following command on your host machine:

sudo killall adb

Issue:

Many Packages don't get downloaded. Example: "error: Download 'NVIDIA Container Runtime with Docker integration (Beta)' failure"

Workaround:

Go to Software and updates and tick on links having Nvidia.

Also subsequently on the jetson try running the following commands

sudo apt update
sudo apt install nvidia-jetpack

Additional Notes:

  • These known issues and workarounds are provided to assist with a smoother installation process.

  • Ensure to follow the instructions carefully to avoid encountering the mentioned issues.


Instructions for manual mode will be updated soon. Please refer to the documentation at for further Reference.

NVIDIA SDK Manager Docker Containers
NVIDIA SDK Manager