LEO PROJECT :-->

BRIEF INTRODUCTION :

This project is based of autonomous path planning and motion of Leo rover. We have taken help of openCV to get the structure of our path ahead and our leo itself decides whether to straight, left or right !

A glimpse of Leo Rover.

INSTALLATION :

This installation procedure assumes that you have installed ROS Version-1 and Gazebo (preferably version 9).

1. The first step is the installation of the leo_simulator. You have to go to this link and clone the leo simulator github repo. You can also go to official ROS website of leo gazebo.

2. After completing the step 1 of installation of, repeat the customary step of compiling the catkin package and sourcing its devel file by:-
   

                  ~/catkin_ws $ catkin build or catkin_make>
                  ~/catkin_ws $ source devel/setup.bash
   

3. In the third step you will clone our repo and place it in your package and again repeat the customary step of compiling the catkin package and sourcing its devel file as stated above.

4. Now, we will move towards installing Differential Drive plugins. If you are using Melodic distro, then you should have this package inbuilt. But if you are on some other distro or somehow not having it, then you may clone this repo and move ahead.

5. Open a fresh terminal and launch the following code in it.
   

             ~/catkin_ws/src $ roslaunch leo_drive spawn_leo.launch
   

After firing this command the folling should appear on your gazebo.

You got it right ?? If yes then you have successfully completed the installation. And now it is : - >

SHOWTIME :

Since we are fully ready with our installation, lets us make our rover work. Follow the following steps to proceed.

1. Open a fresh terminal with the gazebo simulation running on the other one. Fire Up the following command.

        ~/catkin_ws/src $ rosrun leo_drive controller.py
   

If it is running succssfully it should appears as shown in GIF below.

Now, as shown in GIF file, our Rover is able to move on its path.

FUTURE SCOPES :

It is very well known that Autonomous Driving are suffixed with the power of Machine Learning. Therefore, to make our car fully autonomous we may use Sematic Segmentation and Reinforcement Learning to enhance our Leo_Rover learning.
By fine tuning of parameters, it is very much likely to achieve huge improvements in this project.

1. Autonomous Car using Sematic Segmentation by Zsombor Tóth
2. Autonomous Car using Deep Reinforcement Learning.

TEAM MEMBERS :

AYUSH ISHAN
UTKRISHT SINGH
VARAD VINAYAK PANDEY
CHAHAK JETHANI