Documentation

If you need detailed information on the kit’s contents, or how to setup the hardware, please refer to [Dev Kit Documentation](https://docs.baslerweb.com/embedded-vision/daa4200-30mci-jnano-nvdk).

Please note, that the *daA4200-30mci-JNANO-NVDK* Developer Kit and the *daA4200-30mci-JNANO-NVDK-AIA* AI Solution Kit are two different products, so the installed software packages and their version numbers, as described in the "[Debian Packages](https://docs.baslerweb.com/embedded-vision/daa4200-30mci-jnano-nvdk#debian-packages)" section, may differ.

#daA4200-30mci-JNANO-NVDK-AIA

##Step 1 - Initial Configuration:
* The AI Vision Solution Kit is based on the Jetson Nano DevKit. Please setup the system as described in the [Jetson Nano DevKit documentation](https://docs.baslerweb.com/embedded-vision/daa4200-30mci-jnano-nvdk).

##Step 2 - Network Connection:

* Plug in an ethernet cable to connect the AI Vision Solution Kit with the Internet.

* Open a terminal (**ctrl + alt + t**).

* Enter **ip address list dev eth0 |grep -w inet** to find out the network address of the kit.

![ifconfig_inet_grep_ip][1]

* Remember this address as you will need it later to connect via ssh and to open the Web GUI. We will refer to it as the *Nano IP*.

You can now unplug the monitor connection to the kit, as it is now finally ready to be accessed from your workstation computer.

##Step 3 - Setup the AI Vision Solution Kit Software:
* Create an AWS account if you don't have one. One can create an AWS account for free. The new user has access to AWS free tier service for the first 12 months. For more information, please visit [AWS Free Tier](https://aws.amazon.com/free/?all-free-tier.sort-by=item.additionalFields.SortRank&all-free-tier.sort-order=asc)

* Then sign-in to the AWS console [AWS Sign-in](https://signin.aws.amazon.com/signin?redirect_uri=https%3A%2F%2Fconsole.aws.amazon.com%2Fconsole%2Fhome%3Fnc2%3Dh_ct%26src%3Dheader-signin%26state%3DhashArgs%2523%26isauthcode%3Dtrue&client_id=arn%3Aaws%3Aiam%3A%3A015428540659%3Auser%2Fhomepage&forceMobileApp=0&code_challenge=lNVMQoEruNFawQJDdmypoo89M1jGb8MexjLLjRVdArs&code_challenge_method=SHA-256).

* Visit our [Solution Kit update area](https://imaginghub.com/projects/438-ai-vision-solution-kit-with-aws-cloud-connection#files-section) to check for a new [package file](https://imaginghub.com/uploads/basler-ai-vision-solution-kit_1.4_arm64.deb). Download it to your workstation computer.

* Open a terminal session on your workstation computer and change the terminal's current folder to the download location.

* Enter **scp basler-ai-vision-solution-kit\_([major].[minor])\_arm64.deb  ([username]@[Nano IP]):/tmp/** to transfer the downloaded Solution Kit package file to the Solution Kit.

* Enter **ssh -t ([username]@[Nano IP]) sudo dpkg -i /tmp/basler-ai-vision-solution-kit\_([major].[minor])\_arm64.deb** to install the updated Solution Kit package file.

* Enter **ssh -t ([username]@[Nano IP]) ai-vision-solution-kit** to initially download the necessary docker containers (about 500MB), prepare the environment and to start the AI Vision Solution Kit software.

* The execution of the **ai-vision-solution-kit** script is interactive and will take some time depending mainly on your internet connection.

* Note that you must allow sudo access so the configuration can be done. The script will ask you for credentials, if needed.

* In the process you need to enter your **AWS Access Key Id**  and the **AWS Secret Access Key** to use the cloud connector. They will be stored locally on the Nano for later use.

* Note: If you do not want to do this in the part of the process, you can enter random data and use **ssh -t ([username]@[Nano IP]) ai-vision-solution-kit store_credentials [AWS Access Key Id] [AWS Secret Access Key]** later on to setup the correct AWS credentials. These will be needed to enable the connection to AWS (*see step 7 for more details*).

* How to create a new account - [Create and deactivate an AWS account](https://aws.amazon.com/de/premiumsupport/knowledge-center/create-and-activate-aws-account/).

##Step 4 - Get a Live Image:

* Start a web browser on your workstation computer and run the WebGUI using the Nano IP. Checkout Step 2 if you don't know how to find out the Nano IP address.

* Wait, until you can press the Play button.

![ai_web_page_start][2]
##Step 5 - Camera Configuration:
* With the help of the parameters in the WebGUI, the camera can be configured.
* To configure the camera please move the sliders.
* Note: please lower the resolution to get a better frame rate.

![camera_parameters][3]

##Step 6 - AI Models:

* Pre-trained and optimized models for People Localization, Face Localization, Office Item Detection and Mask Detection are available in the WebGUI for deployment.
* Choose one of these inference-models and hit the apply button to run it.
* To see a result image switch switch from “Live Image” to “Result Image” by using the controls you see when moving the mouse pointer inside the image.
* Note that the initial download and build of a model will take some minutes. As the result will be cached future use is much faster (about 35s).
![WebGUI_pretrained_models][4]

##Step 7 - AWS Connection:

With the Solution Kit, you have the option to send the resulting meta data from your inference-models to the [AWS IoT Core][5]. For more details, please refer to [this AWS documentation][6] to learn how to get started with the AWS IoT Core.

* First make sure that your AWS credentials are entered correctly as described in step 3.
* Access the WebGUI and switch the **send results to AWS** switch to **ON**. If the switch is grayed out and not working check if the credentials are entered correctly.
![slider_grayed_out][7]
* Restart the ai-kit by entering **ai-vision-solution-kit restart** in the terminal on the Jetson Nano.
* Reload the WebGUI, load a model as described in step 6.
* Access your AWS-Console, select the IoT-Core service and go to **Test**.
* Subscribe to **aikit/Results**. You should now see the results, sent to AWS.
![aws_test][8]

###From here on you might want to do one of the following:
* [How to create, train and deploy a Machine Learning Model?](https://imaginghub.com/projects/441-how-to-create-train-and-deploy-a-machine-learning-model)
* [How to optimize a Machine Learning Model][9]
* [FAQ][10]

[1]: https://imaginghub.com/uploads/editor/images/2020/5e922d16-7e13-4a2a-ac12-59b8f0dd4769.png
  [2]: https://imaginghub.com/uploads/editor/images/2020/1c47e708-54d8-4424-8b91-8f4511f1a5ca.png
  [3]: https://imaginghub.com/uploads/editor/images/2020/3a3de5d1-8df5-405d-895c-82b70feb59f9.png
  [4]: https://imaginghub.com/uploads/editor/images/2020/eb32e052-0791-4bf2-9803-5774a1947bed.png
  [5]: https://docs.aws.amazon.com/iot/index.html
  [6]: https://docs.aws.amazon.com/iot/latest/developerguide/iot-gs.html
  [7]: https://imaginghub.com/uploads/editor/images/2020/f9f81ba2-b2f0-4574-81c3-640dfc88a344.png
  [8]: https://imaginghub.com/uploads/editor/images/2020/6ece9cc3-1645-41ee-a8a4-94012c7bc22c.png
  [9]: https://imaginghub.com/projects/442-how-to-optimize-a-machine-learning-model
  [10]: https://imaginghub.com/projects/445-faq

If you need detailed information on the kit’s contents, or how to setup the hardware, please refer to Dev Kit Documentation.

Please note, that the daA4200-30mci-JNANO-NVDK Developer Kit and the daA4200-30mci-JNANO-NVDK-AIA AI Solution Kit are two different products, so the installed software packages and their version numbers, as described in the "Debian Packages" section, may differ.

daA4200-30mci-JNANO-NVDK-AIA

Step 1 - Initial Configuration:

The AI Vision Solution Kit is based on the Jetson Nano DevKit. Please setup the system as described in the Jetson Nano DevKit documentation.

Step 2 - Network Connection:

Plug in an ethernet cable to connect the AI Vision Solution Kit with the Internet.
Open a terminal (ctrl + alt + t).
Enter ip address list dev eth0 |grep -w inet to find out the network address of the kit.

ifconfig_inet_grep_ip

Remember this address as you will need it later to connect via ssh and to open the Web GUI. We will refer to it as the Nano IP.

You can now unplug the monitor connection to the kit, as it is now finally ready to be accessed from your workstation computer.

Step 3 - Setup the AI Vision Solution Kit Software:

Create an AWS account if you don't have one. One can create an AWS account for free. The new user has access to AWS free tier service for the first 12 months. For more information, please visit AWS Free Tier
Then sign-in to the AWS console AWS Sign-in.
Visit our Solution Kit update area to check for a new package file. Download it to your workstation computer.
Open a terminal session on your workstation computer and change the terminal's current folder to the download location.
Enter scp basler-ai-vision-solution-kit_([major].[minor])_arm64.deb ([username]@[Nano IP]):/tmp/ to transfer the downloaded Solution Kit package file to the Solution Kit.
Enter ssh -t ([username]@[Nano IP]) sudo dpkg -i /tmp/basler-ai-vision-solution-kit_([major].[minor])_arm64.deb to install the updated Solution Kit package file.
Enter ssh -t ([username]@[Nano IP]) ai-vision-solution-kit to initially download the necessary docker containers (about 500MB), prepare the environment and to start the AI Vision Solution Kit software.
The execution of the ai-vision-solution-kit script is interactive and will take some time depending mainly on your internet connection.
Note that you must allow sudo access so the configuration can be done. The script will ask you for credentials, if needed.
In the process you need to enter your AWS Access Key Id and the AWS Secret Access Key to use the cloud connector. They will be stored locally on the Nano for later use.
Note: If you do not want to do this in the part of the process, you can enter random data and use ssh -t ([username]@[Nano IP]) ai-vision-solution-kit store_credentials [AWS Access Key Id] [AWS Secret Access Key] later on to setup the correct AWS credentials. These will be needed to enable the connection to AWS (see step 7 for more details).
How to create a new account - Create and deactivate an AWS account.

Step 4 - Get a Live Image:

Start a web browser on your workstation computer and run the WebGUI using the Nano IP. Checkout Step 2 if you don't know how to find out the Nano IP address.
Wait, until you can press the Play button.

ai_web_page_start

Step 5 - Camera Configuration:

With the help of the parameters in the WebGUI, the camera can be configured.
To configure the camera please move the sliders.
Note: please lower the resolution to get a better frame rate.

camera_parameters

Step 6 - AI Models:

Pre-trained and optimized models for People Localization, Face Localization, Office Item Detection and Mask Detection are available in the WebGUI for deployment.
Choose one of these inference-models and hit the apply button to run it.
To see a result image switch switch from “Live Image” to “Result Image” by using the controls you see when moving the mouse pointer inside the image.
Note that the initial download and build of a model will take some minutes. As the result will be cached future use is much faster (about 35s).

Step 7 - AWS Connection:

With the Solution Kit, you have the option to send the resulting meta data from your inference-models to the AWS IoT Core. For more details, please refer to this AWS documentation to learn how to get started with the AWS IoT Core.

First make sure that your AWS credentials are entered correctly as described in step 3.
Access the WebGUI and switch the send results to AWS switch to ON. If the switch is grayed out and not working check if the credentials are entered correctly.
Restart the ai-kit by entering ai-vision-solution-kit restart in the terminal on the Jetson Nano.
Reload the WebGUI, load a model as described in step 6.
Access your AWS-Console, select the IoT-Core service and go to Test.
Subscribe to aikit/Results. You should now see the results, sent to AWS.