Warehouse Drone

Navigation and Control Systems

[Code]

This project focuses on developing an autonomous Warehouse Drone for the e-Yantra 2023-24 competition, leveraging advanced techniques in navigation, control systems, and simulation. The drone is designed to autonomously navigate a confined warehouse environment, localize itself, and perform tasks efficiently.

Key Features:

  • Navigation: Implemented A* algorithm for optimal pathfinding in cluttered environments.
  • Control Systems: Used PID tuning for precise control over drone movements, ensuring smooth and accurate navigation.
  • Localization: Applied computer vision techniques to accurately determine the drone’s position within the environment.
  • Simulation: Developed and tested in Gazebo Ignition, providing a realistic simulation environment to fine-tune performance.
Left: Simulation environment in Gazebo Ignition. Right: A* pathfinding visualization during navigation.

Progress and Results

  • Successfully cleared Stage 1 of the competition.
  • Currently ranked 15th in the ongoing leaderboard.
  • Demonstrated robust control and navigation capabilities under simulation conditions.

Workflow and Methodology

  1. Navigation Algorithm:
    • Implemented A* for computing the shortest path between start and goal locations, optimizing for obstacles in the environment.
  2. Control Systems:
    • Tuned PID controllers for maintaining stable and responsive drone movement.
  3. Localization:
    • Leveraged computer vision techniques such as feature matching for precise localization.
  4. Simulation and Testing:
    • Utilized Gazebo Ignition for a realistic warehouse simulation environment to iterate and refine drone behavior.

Future Goals

  • Advance to the next competition stage by further improving navigation efficiency and robustness.
  • Transition from simulation to hardware deployment for real-world testing.