Skip to content

BUILD.md

Latest commit

 

History

History
183 lines (118 loc) · 4.58 KB

BUILD.md

File metadata and controls

183 lines (118 loc) · 4.58 KB

NVIDIA Jetson Nano setup manual

Manual

Install the operating system

Follow the instructions at https://developer.nvidia.com/embedded/learn/get-started-jetson-nano-devkit#write to install the operating system to the Jetson Nano. Make sure to finish the installation process, e.g. setting the timezone and creating a user account.

Upgrade and cleanup

Upgrade the system and automatically remove unnecessary packages

sudo apt update && sudo apt upgrade -y && sudo apt autoremove -y

Install prerequisites

Before starting the installation process, please ensure that git and python3-pip are installed.

sudo apt install nano git python3-pip

Optional: Strip the operating system

Because the standard Jetson Nano operating system comes with a lot of tools which won't be used, we can save valuable space and resources by removing them.

sudo apt purge snapd lx* nautilus* openbox* printer* rhythmbox* gnome* lightdm* xubuntu* xscreensaver* xfce* lxde* x2go* word* thunderbird libreoffice* chromium-* vim* docker* libvisionworks* ubuntu-wallpapers-bionic light-themes adwaita-icon-theme -y && sudo apt autoremove -y

This command will strip the operating system of most unnecessary packages. Please read the command before executing to avoid removing necessary packages.

Installation of the project

Clone the project from the GitHub repository.

git clone https://github.com/sieuwe1/Autonomous-Ai-drone-scripts.git

Project packages

Jetson Inference

Start by installing the required packages for the installation process.

sudo apt install cmake libpython3-dev python3-numpy

Execute the following commands in sequence to install Jetson Inference.

During the installation, make sure to deselect all models and finally select the SSD MobileNet V2 model. After the download has completed, please also install PyTorch.

git clone --recursive https://github.com/dusty-nv/jetson-inference
cd jetson-inference
mkdir build
cd build
cmake ../
make -j$(nproc)
sudo make install
sudo ldconfig

Source: https://github.com/dusty-nv/jetson-inference/blob/master/docs/building-repo-2.md

Remaining packages
python3 -m pip install --upgrade pip
python3 -m pip install scikit-build cmake opencv-python numpy dronekit pyserial keyboard simple-pid

Utilities

X11 forwarding

For viewing sensors, media and other things with a graphical user interface, we can use X11 forwarding to forward the interfaces over SSH.

Enable X11 forwarding in the sshd configuration file.

sudo nano /etc/ssh/sshd_config

Look for a parameter called X11Forwarding, ensure that the value is set to yes. It should look like this.

X11Forwarding yes

To connect to the Jetson Nano with X11 support, use the following command:

ssh -X -Y USERNAME@IP
Viewing videos

A common use case for X11 within the drone project is to watch back debugging videos. To review debugging videos from the drone via X11, you can use the mplayer package, which is a lightweight media player.

Install mplayer

sudo apt install mplayer

Example of mplayer usage

mplayer video.mp4

Simulator environment installation

Clone the simulator from ArduPilot from the GitHub repository.

git clone https://github.com/ArduPilot/ardupilot --recursive
cd ardupilot

Source: https://ardupilot.org/dev/docs/building-setup-linux.html#building-setup-linux

Environment prerequisites

Before executing the prerequisite installation script from ArduPilot, please install some additional packages.

sudo apt install libsdl2-dev libsdl2-image-dev libsdl2-mixer-dev libsdl2-ttf-dev libfreetype6-dev python3-setuptools python3-dev python3 libportmidi-dev

After that, install all prerequisites using the tool from ArduPilot by executing the following command:

Tools/environment_install/install-prereqs-ubuntu.sh -y

Reload the path (log-out and log-in to make permanent).

. ~/.profile

Build

Before building, ensure that the project is configured.

./waf configure

After the project is configured, build the project.

./waf clean
./waf build

Startup

Ensure that the current directory is the scripts directory.

cd Tools/scripts

Start the simulator for a ArduCopter

sim_vehicle.py -v ArduCopter --map --console

Cleanup

Finish by removing all unused packages.

sudo apt autoremove -y