CarND-Capstone project's Docker container

CarND-Capstone is the last project of the Udacity self-driving car NanoDegree. It's composed of a simulator (running on your local machine) and the brain of the car (running in this container, including compilation and debugging).

The container connects to the port 4567 of the host machine, that's the port the simulator listens on.

See https://github.com/udacity/CarND-Capstone

Supported OSes

Linux and MacOS are supported. ( CUDA only tested on Linux so far)

CUDA support

Install nvidia-docker from here: https://github.com/NVIDIA/nvidia-docker

MacOS

I recommend to install Docker for MacOS instead of managing your own VM (port forwarding and file sharing is automated for you).

Installation

1.Go to your project directory

cd /home/user/capstone_source

2.Download the script into your project

wget https://raw.githubusercontent.com/SiliconCar/CarND-Capstone-Docker/master/utils/run_cuda.sh && chmod u+x run_cuda.sh

Usage

1.Start the script

./run_cuda.sh

2.You're now within the container and the current directory contains the source code of your project. You can compile and run your project:

cd ros
catkin_make
source devel/setup.sh
roslaunch launch/styx.launch

Aliases

Here are some of the cool aliases that are defined in the container

udacity_make #Compiles the project
udacity_run #Executes the project

Multiple bash sessions

Simply open a new terminal and run the script again, if your container is currently running it will attach to it. Therefore, you can have one terminal session to run the project and another one to query ROS's topic or what have you.