This project implements an autonomous distributed sorting system that orchestrates collaboration between a Braccio robot arm (for object detection & manipulation) and a DJI RoboMaster (for mobile transport). The system leverages computer vision and parallel processing to identify, classify, and sort objects efficiently.

Developed as a Master's Team Project for the Summer Semester 2025 at Technische Hochschule Ingolstadt.

• Dual-Robot Coordination: Synchronized operations between a stationary manipulator (Braccio) and a mobile unit (DJI RoboMaster).
• Computer Vision: Real-time object detection and color classification using custom models.
• Parallel Processing: Multithreaded architecture ensuring non-blocking operations for both robots.
• Modular Architecture: Decoupled modules for vision, movement, and control logic.

The system is built on a modular Python architecture located in the sorting_system package:

.
├── sorting_system/
│   ├── braccio/           # Logic for the Braccio robotic arm
│   │   ├── color_detection.py
│   │   └── pick_place.py
│   ├── dji/               # Logic for the DJI RoboMaster
│   │   ├── dji_movement.py
│   │   └── dji_camera.py
│   └── main.py            # Central controller & entry point
├── models/                # YOLOv5 inference models
└── requirements.txt       # Dependency definitions

• Hardware:
  • Arduino Braccio Robot Arm
  • DJI RoboMaster S1/EP
• Software:
  • Python 3.8+
  • RoboMaster SDK

1. Clone the Repository

   git clone https://github.com/YOUR_USERNAME/sorting-system.git
   cd sorting-system

2. Install Dependencies

   pip install -r requirements.txt

3. Model Setup Ensure your YOLOv5 model (yolov5n.pt) is placed in the models/ directory.

1. Connect Hardware: Ensure both the Braccio (via USB/Serial) and DJI RoboMaster (via Wi-Fi/Direct) are connected.
2. Run the System:

   python sorting_system/main.py

3. Operation: The system will initialize the camera feeds, start the detection threads, and await object placement for sorting.

• Advanced Object Detection: Training a custom YOLO dataset for specific industrial component recognition.
• Safety Mechanisms: Integrating an emergency stop button and collision detection algorithms.
• Dynamic Path Planning: Implementing obstacle avoidance for the DJI RoboMaster in dynamic environments.

• Team Project - Initial Work - Technische Hochschule Ingolstadt

This project is licensed under the MIT License - see the LICENSE file for details.