abest0/aws-robomaker-jetbot-detect

Let's use AWS RoboMaker to build and deploy ROS applications on a jetbot for autonomous navigation and more...

AWS Robomaker + SparkFun NVidia JetBot ROS Application

This sample ROS1 Melodic application leverages AWS SageMaker and AWS IoT

Setup your AWS RoboMaker Development and Build Environment

First, sign into the AWS Management Console and click on AWS RoboMaker. Create a new development environment using the default settings.

1. Clone this repository into your Cloud9 environment.

       git clone https://github.com/jerwallace/aws-nvidia-sample-robomaker-dino-detect
       cd aws-nvidia-sample-robomaker-dino-detect
   

2. Before we get started, you will need an S3 bucket to store your assets. Run the following command in Cloud9

       aws s3 mb <YOUR_BUCKET_NAME>
   

3. Copy the roboMakerSettings.json.template file to roboMakerSettings.json

    cp roboMakerSettings.json.template roboMakerSettings.json
    # Edit roboMakerSettings.json and replace where prompts exist

1. Next, we need to create a set of IoT credentials, so our application can communicate with AWS IoT. Run the following shell script to generate these:

       sudo chmod -R +x scripts/*.sh
       . scripts/generate_certs.sh
   

2. Now, we are ready to create our cross-compiler docker image. First, open the file cross-compiler/Dockerfile in the Cloud9 editor window. At the top of the file you will see a few arguments.
3. In Cloud9, open a seperate terminal window as this command will take a while.

       cd cross-jetson-nano/
       docker build ./
   

Congratulations! You are ready to start building ROS Melodic applications.