Robot Code

Code that runs on the robot itself. Creates a Racecar object with camera for it to be exploited by this code.

Pre-requisites

• Jetson Nano AI Kit 4GB - Jetpack 4.5.1
• Python 3.6
• Git

Installation

If not done already, install the needed python dependencies on the robot. You can do that by running the following command:

pip install -r requirements.txt

Running the code

Note

This repository should be on the Jetson Nano directly to help with file transfer and development.

• First, you need to replace the WebScoket server's IP in the web_socket_client.py and kpi_plotter_ws.py files with the IP on which your server runs.
  Then, run the WebSocket solution code, simply run this command:

  python3 web_socket_client.py

• To run the TCP Tailscale solution code, simply run this command:

  python3 direct_tcp.py

Note

After starting the script, it might take a while for the camera and controls to fully initialize. Please be patient as this can take up to 2 minutes.

There should be a message in the logs letting you know that the script is ready when its done initializing.

KPIs (Key Performance Indicators)

The present code has telemetry that is taken during runtime to evaluate the performance of the different solutions we have.

After each run, it will generate .csv files with all the data used to create plots (useful for modifying data to remove outliers).

It will also create 2 .png files with 3 plots each containing different data sets.

---

Made with care by @Funnyadd, @ImprovUser, @cjayneb and @RaphaelCamara ❤️