Skip to content

Driving Your Donkey Car and Collecting Data

Imagine your donkey car is a small, smart robot car that you can control with a remote. To get it moving and learn how to switch between different driving modes, you'll start with a simple command on your computer. Here's how you do it:

  1. Start the Car: Open your computer's command line tool, type python manage.py drive --js, and press Enter. This wakes up your donkey car and gets it ready to follow your commands.

    Drive the Donkey Car
    python manage.py drive --js

  2. Understanding the Controls: Right after you run the command, you'll see instructions pop up on your screen. These tell you which buttons on your controller do what. It's like learning the controls of a new video game. Make sure to jot these down or take a picture with your phone - you'll need to refer back to them!

  3. Switching Driving Modes: Your donkey car has a cool feature - it can drive in different ways! There's a special button (usually the start button) that lets you switch between:

    1. Fully Remote-Controlled Mode: You control everything, just like driving an RC car.

    2. Fully Autonomous Mode: The car drives all by itself, making its own decisions on turning and speed.

    3. We'll focus on these two modes. If your controller doesn't seem to respond, hitting the start button is a good first troubleshooting step.

Collecting Data for Your Car to Learn From

Now that your car can move around, it's time to teach it how to drive on its own. This is done by collecting data - basically, you drive the car around and it remembers how you did it. Here's how to gather good learning material for your car:

  1. Drive Around: You'll need to drive your car around the track in both directions. Aim for about 10 laps each way. This gives your car a variety of examples to learn from.

  2. It's Okay to Make Mistakes: Try to keep the car within the track lines, but don't worry about staying perfectly centered all the time. In fact, it's good for your car to see how to recover from being off-center. This helps it learn to correct itself and makes it smarter at handling different situations.

Remember, the goal isn't to collect flawless data but to give your car a rich learning experience, full of different scenarios and recoveries. This way, your car becomes more adaptable and can handle the track like a pro, even when things don't go exactly as planned.