Skip to content
/ nxa Public

An self-driving RC car based on ROS, Raspberry Pi, TensorFlow, Keras, CNN, and Donkey® Car API

Notifications You must be signed in to change notification settings

hchen98/nxa

Repository files navigation

nxa

An RC car based on ROS, Raspberry Pi, Keras, and Donkey® Car API

Demo:

IMAGE ALT TEXT HERE


Requirements:

  • Python 3.x
  • Raspberry Pi 3/ 4 (recommended and used in this project) or Nvidia Jetson Nano
  • Servo Driver PCA 9685
  • Magnet Car or alternative
  • ROS/ Raspbian Lite
  • 3D printed parts

Brief View of Hardware Setup:


Collecting Data:

  1. Collbrate with Pi:
$ donkey calibrate --channel x/x
  1. Drive the Pi:
$ python drive.py
  1. View the drive control panel on host computer: Opening your web broswer and typing your Pi's IP address and port 8887 e.g., 127.0.0.1:8887

To get start with training model and drive:

  1. Trand your Keras Model
$ python train.py [tub directory]
  1. Drive with your trained model
$ python train.py [tub directory]

NOTE: Donkey® Car project provides Unity simulator for training. Chech out here!

Releases

No releases published

Packages

No packages published

Languages