Autonomous Package Delivery Robot

An autonomous mail delivery system built on the TurtleBot 3 Waffle Pi, combining Bayesian Localization, PID control, and ROS for precise navigation and reliable delivery in a simulated hallway environment.

---

Features

1. Line Following:

   • Implemented using a finely-tuned PID controller to ensure smooth and precise navigation along white tape paths.
   • Dynamically adjusts angular velocity to minimize overshoot and improve tracking accuracy.

2. Office Recognition:

   • Detects delivery locations using HSV color thresholds.
   • Adapted to handle challenging lighting conditions by distinguishing colors based on hue and saturation.

3. Bayesian Localization:

   • Employs a BayesLoc class to predict the robot's state and update probabilities based on observed colors.
   • Achieves convergence on the robot's position with >70% confidence.

4. Mail Delivery:

   • Automates delivery maneuvers, including stopping, 90-degree turns, and simulated package drops.
   • Executes actions only when localization confidence exceeds a threshold.

5. Real-Time Control:

   • Utilizes ROS for seamless integration of hardware and software, including sensor processing and motor control.

System Architecture

Hardware

• Platform: TurtleBot 3 Waffle Pi
  • Microcontroller: OpenCR1.0 (ARM Cortex-M7)
  • High-Level Processor: Raspberry Pi 3
  • Motors: Dynamixel XM430-W210 actuators
• Sensors:
  • Raspberry Pi Camera Module v2.1 (for line detection and office recognition)

Software

• Programming Languages: Python (main logic), ROS (middleware)
• Control Algorithm: PID controller
• Localization Algorithm: Bayesian Localization

Documentation

• Full report: Project Documentation
• Demo video: YouTube Link

---

License

This project is licensed under the MIT License.