# Installation

Installation instructions for various operating systems are available below. Presently, these are the only supported operating systems for the Scout Agent. After completing the installation you will need to register your license(s) and then select a data destination, you can choose between four data destination options depending on how you wish to use the Scout Agent.

{% tabs %}
{% tab title="Windows" %}

1. For the most recent version, download the [Rekor Scout Windows Installer](https://deb.openalpr.com/windows-agent/openalpr-agent-latest.exe). For previous versions, the download is available on the [Release Notes](https://docs.rekor.ai/scout/agent/agent-release-notes) page.
2. Right-click the .exe file, run as administrator, and follow the install wizard.
   {% endtab %}

{% tab title="Ubuntu Linux" %}
Installation on Linux requires basic knowledge of using the terminal. If you do not already have Linux installed as an operating system, please complete these steps before continuing:

1. Download the [Ubuntu 20.04 64-bit install image](https://releases.ubuntu.com/20.04/) and copy it to a bootable flash drive.
2. Follow this [installation guide](http://www.ubuntu.com/download/desktop/install-ubuntu-desktop).

If you already have Linux installed:

1. Open a new terminal
2. Ensure you have curl installed by running the following command from the terminal:

   ```sh
   sudo apt-get update && sudo apt-get install -y curl
   ```
3. Run our installer script and choose `install_agent`

   ```shell
   sudo apt update && sudo apt install -y curl
   bash <(curl -s https://deb.openalpr.com/install) # select "install_agent"
   ```
4. Select yes/no when prompted to connect the Agent to your cloud account
5. If you are using a GUI, you can configure the agent using the alprdconfig utility by running the following command:
6. If your server does not have a GUI and you wish to connect your software to a Scout web server, do the following:
   1. `alprlink-register -w https://[ip_address_of_web_server]`
   2. Enter the e-mail address and password for your Scout Cloud account, or the one you used when installing the on-premise webserver.
      {% endtab %}

{% tab title="Docker" %}

1. Open a new terminal
2. Pull the latest OpenALPR Docker image

   ```docker
   docker pull openalpr/agent:latest
   ```
3. Enable the license for the agent

   ```docker
   docker run -P -v openalpr-vol1-config:/etc/openalpr/ -v openalpr-vol1-images:/var/lib/openalpr/ -it openalpr/agent openalpr-licenseregister
   ```
4. Configure it to point to a Scout web server (default is `https://cloud.openalpr.com`)

   ```docker
   docker run -P -v openalpr-vol1-config:/etc/openalpr/ -v openalpr-vol1-images:/var/lib/openalpr/ -it openalpr/agent alprlink-register
   ```
5. Start the container

   ```docker
   docker run --restart always -d --cap-add SYS_NICE -P -v openalpr-vol1-config:/etc/openalpr/ -v openalpr-vol1-images:/var/lib/openalpr/ -it openalpr/agent
   ```

{% endtab %}

{% tab title="Jetson" %}
The Scout Agent can be installed on NVIDIA Jetson devices such as the Nano, TX-1, and TX-2. These are excellent options for an edge device due to the power-efficient, embedded GPU. Native binaries are available for JetPack versions 4.3-4.6. The recommended package for other JetPack versions is the Jetson Docker image.

* **JetPack 4.3.0-4.6.x** : follow Linux install instructions
* **JetPack 5.1.X:** pull the following Docker image:

```
docker pull openalpr/agent:5.0.0-jetson51-sdk
```

Note that both native and Docker installations will automatically include the additional GPU packages. Therefore, you will not need to reference the GPU acceleration section of the documentation. When installing via Docker, make sure you have [NVIDIA Docker Runtime](https://github.com/NVIDIA/nvidia-docker) available so that the container can access the host GPU. You will also need to modify any Docker run commands.

```docker
docker run --runtime nvidia -e NVIDIA_DRIVER_CAPABILITIES=compute,video [other_flags] [image_tag] [command]
```

{% endtab %}
{% endtabs %}

{% embed url="<https://vimeo.com/461857506/9900fe8cb0?share=copy>" %}
