leandroveronezi/dart-self-driving-car

This project is a simple example of a self-driving car controlled by an artificial neural network, implemented in Dart.

Self-driving car in Dart

This project is a simple example of a self-driving car controlled by an artificial neural network, implemented in Dart.

The goal is to learn about neural networks and how to use them to control a virtual car in a game environment.

Inspiration:

The project is inspired by the Self-driving Car (no libraries) JavaScript course.

What you need:

• Dart SDK installed: https://dart.dev/
• Git installed: https://git-scm.com/

Getting started:

1. Clone the repository:

   git clone https://github.com/leandroveronezi/dart-self-driving-car.git

2. Access the project folder:

   cd dart-self-driving-car

3. Activate webdev:

   dart pub global activate webdev

4. Start the server:

   webdev serve

5. Access the project:

   http://localhost:8080
   

How it works:

• The neural network has 9 inputs, representing the distance to the obstacles in front.
• The neural network has 4 outputs, representing the vehicle control.
• The car is controlled by the neural network, which is trained to minimize the error relative to a goal. The goal is for the car to follow a track without colliding with walls or other cars.

Developing:

• Explore the code to understand how the neural network is implemented and how the car is controlled.
• Experiment with different neural network configurations to see how the car's performance changes.
• Implement new features, such as obstacles or different tracks.

Resources:

• Dart: The programming language.
• Webdev: Tool for web development in Dart.

Notes:

• The code is for educational purposes only.

Enjoy learning!

Bonus: Play the game