ROS-based Robot Fleet Navigation System (In Development)

This repository contains the code developed for the Flipkart Grid 3.0 robotics competition, where we achieved the 2nd Runner-up position. The challenge required participants to create a package-sorting fleet management system using only a single overhead camera as the sole sensing element. No cameras or LiDAR sensors were allowed on the robots themselves.

Project Status

• Core Functionality: The competition code is fully functional and implements a centralized fleet navigation system using ROS and OpenCV, relying solely on the overhead camera for perception.
• Simulation: Development is ongoing for a simulation environment using Gazebo.

Getting Started

Prerequisites

Ensure you have the following installed on your system:

• ROS Noetic
• Gazebo 9

Installation & Setup

1. Clone the repository:

    git clone https://github.com/nsk-05/ros_atlas

2. Move to project folder

    cd ros_atlas
   

3. Launch the simulations and nodes

    roslaunch bot_gazebo grid.launch
   
    roslaunch bot_gazebo map_server.launch
   
    rosrun bot_vision goal_manager.py
   

4. Spawn delivery package in robot's package holder

    rosrun bot_package_manager package_spawner.py
   

Multi-agent conflict based search(MACBS) visualisation