Quick Start

Note: for Mac users, please dual boot as Windows 10.

  1. Download Simulator distribution for your OS:

    For competition, please download Berkeley Major Map!

  2. Fork the Repo

    • Please click the Fork button on the upper right corner. Note you would need to sign in to your GitHub account first, and your account will be referred to below as YOURUSERNAME.

  3. clone the repo

    • git clone --recursive https://github.com/YOURUSERNAME/ROAR.git
      • replace YOURUSERNAME with your github username.

  4. Create virtual environment with python3.8

    • To run Berkeley_Major_Map and evGrandPrix_Map, create python3.8 environment:
      • conda create -n ROAR_major python=3.8
      • conda activate ROAR_major
    • To run Berkeley_Minor_Map and Easy_Map, create python3.7 environment:
      • conda create -n ROAR_minor python=3.7
      • conda activate ROAR_minor

  5. Install Dependency

    • General Dependency
      • pip install -r requirements.txt in the ROAR folder
      • If any specific package version cannot be installed, install a downgrade version or install lastest avaiable version:
        • pip install -r requirements_general.txt

  6. To run simulator:

    • open the carla map
    • run simulation script runner_sim.py:
      • python runner_sim.py


** To run on a physcial car (The jetson code is out of date and out of support, it is included here just for reference):

  • For actual vehicle wired to your computer
    • cd ROAR_Jetson
    • pip install -r requirements.txt
  • For actual vehicle running on Jetson Nano
    • cd ROAR_Jetson
    • sudo apt-get install python-dev libsdl1.2-dev libsdl-image1.2-dev libsdl-mixer1.2-dev libsdl-ttf2.0-dev libsdl1.2-dev libsmpeg-dev python-numpy subversion libportmidi-dev ffmpeg libswscale-dev libavformat-dev libavcodec-dev libfreetype6-dev
    • pip3 install -r requirements_jetson_nano.txt
  • Run script:
    • python runner_jetson.py or python3 runner_jetson.py