Envision

Envision is a scenario generation language for Gazebo. Envision is designed to make creating world files in Gazebo simple and painless. Instead of trying to manually create worlds in Gazebo, Envision will allow you to choose a premade scene and add models to it using an input file.

Table of Contents

• Software Requirements
• Installation
  1. Gazebo
  2. Envision
  3. Model Library
• Dependencies
• Directories
• How To Run (Example)
• After Running
• Documentation
• Authors

Software Requirements

• Ubuntu 18.04 or 20.04
• Python >= 3.8 (not tested on older versions)

Installation

1. Gazebo

Installing Gazebo

$ curl -sSL http://get.gazebosim.org | sh

2. Envision

Download Envision from GitHub...

$ git clone https://github.com/Vikings1028/envision.git

Dependencies

pip3 install html-testRunner argparse

Directories

Name	Contents
docs	Class and project documentation
inputfiles	Input files that will be analyzed by envision
logs	Log files created at runtime
src	Source files
test	Test code and unit tests
worlds	Created .world files

How To Run

To run Envision, you will create an input file and place it in the inputfiles directory. By default, a sample file will be included called sample1.txt that will guide you through the process of entering data. The InputReadMe file includes all of the possible models and scenes.

Following the creation of the input file, the run.py file will be used. The follwing switches are available:

-h or --help: Display possible switches

-f or --file: Indicates the text file to be used

-t or --test: Testing mode

-w or --world: Output .world file name

-o or --open: Open .world file in Gazebo after running

Change directory to envision...

$ cd PATH_TO_ENVISION/

Run Envision with just a file...

$ python3 run.py -f sample1.txt

Run Envision with a file and world name...

$ python3 run.py -f sample1.txt -w newWorld

Run Envision with a file and open in Gazebo...

$ python3 run.py -f sample1.txt -o

Run Envision in test mode...

$ python3 run.py -t

After Running

Envision will output a .world file inside of the worlds directory. If the world name was specified, the world file will be named that. If the world name was not specified, the world file will be called user_defined.world.

The world file can then be opened in gazebo by calling Gazebo along with the world file.

If trying to output user_defined.world in Gazebo, you would type

$ gazebo user_defined.world

Authors

• Jake Summerville
• Safwan Elmadani