Data

Tools

Indexes

Applications

Experiments

Awesome

An open API service indexing awesome lists of open source software.

https://github.com/hasanulmukit/ai-smart-factory-optimization

An AI/ML system designed to optimize smart factory operations by streamlining production lines, reducing waste, and automating material recycling.
https://github.com/hasanulmukit/ai-smart-factory-optimization

artificial-intelligence machine-learning recycling streamlit waste-classification

Last synced: 11 months ago
JSON representation

An AI/ML system designed to optimize smart factory operations by streamlining production lines, reducing waste, and automating material recycling.

Awesome Lists containing this project

README 

          
## Smart Factory Optimization with Automated Material Recycling



**Repository Overview**



This project implements an AI/ML system designed to optimize smart factory operations by streamlining production lines, reducing waste, and automating material recycling. Leveraging state-of-the-art deep learning models and advanced data processing tools, this repository guides you through every step—from data collection and preprocessing to model training, inference, and dashboard deployment using Streamlit.



---



## Table of Contents



* [Features](#features)

* [System Architecture](#system-architecture)

* [Prerequisites](#prerequisites)

* [Installation](#installation)

* [Project Structure](#project-structure)

* [Data Preparation](#data-preparation)

* [Model Training](#model-training)

* [Inference Pipeline](#inference-pipeline)

* [Streamlit Dashboard](#streamlit-dashboard)

* [Advanced Data Processing with DeepSeek](#advanced-data-processing-with-deepseek)

* [Testing and Optimization](#testing-and-optimization)

* [Deployment](#deployment)

* [Contributing](#contributing)

* [License](#license)

* [Acknowledgments](#acknowledgments)



---



## Features



* **Image-based Waste Classification:** Uses transfer learning on EfficientNet to classify recyclable vs. non-recyclable materials.

* **Data Augmentation:** On-the-fly augmentation to improve model robustness.

* **Automated Inference Pipeline:** Real-time image ingestion and classification.

* **Interactive Dashboard:** Streamlit app for uploading images, visualizing predictions, and monitoring performance metrics.

* **DeepSeek Integration:** Advanced data processing and report generation on production metrics.

* **Extensible Design:** Easily swap model architectures or integrate object detection frameworks like YOLO.



---



## System Architecture



```plaintext

+---------------------+       +----------------------+       +--------------------+

| Production Line Cam |  -->  | Inference Pipeline   |  -->  | Automated Sorting  |

+---------------------+       +----------------------+       +--------------------+

         |                                                          ^

         v                                                          |

+----------------------+       +----------------------+                |

| Data Collection      |       | Model Training       |-----------------+

+----------------------+       +----------------------+                |

         |                                                          |

         v                                                          v

+----------------------+       +----------------------+       +----------------+

| DeepSeek Reporting   |<--    | Streamlit Dashboard  |       | Visualization  |

+----------------------+       +----------------------+       +----------------+

```



---



## Prerequisites



* Python 3.8+

* Kaggle or Google Colab for model training (with GPU runtime)

* VS Code for local development

* Streamlit for dashboard

* Git for version control



---



## Installation



1. Clone the repository:



   ```bash

   git clone https://github.com/yourusername/smart-factory-recycling.git

   cd smart-factory-recycling

   ```



2. (Optional) Create and activate a virtual environment:



   ```bash

   python3 -m venv env

   source env/bin/activate        # macOS/Linux

   env\Scripts\activate.bat     # Windows

   ```



3. Install dependencies:



   ```bash

   pip install -r requirements.txt

   ```



---



## Project Structure



```plaintext

smart-factory-recycling/

├── dataset/                       # Raw and processed images

├── data_preprocessing.py         # Data loading & augmentation

├── model_training.py             # Model definition & training loop

├── inference.py                  # Inference utilities

├── dashboard.py                  # Streamlit dashboard

├── deep_insights.py              # DeepSeek integration scripts

├── requirements.txt              # Python dependencies

├── best_model.h5                 # Saved best model weights

└── README.md                     # Project documentation

```



---



## Data Preparation



1. **Organize Dataset:** Place images in `dataset//` directories (e.g., `dataset/recyclable/`, `dataset/non_recyclable/`).

2. **Run Preprocessing:**



   ```bash

   python data_preprocessing.py

   ```

3. **Inspect Augmentations:** The script visualizes sample augmentations to verify correctness.



---



## Model Training



1. **Configure Hyperparameters:** Modify constants in `model_training.py` (e.g., `IMG_HEIGHT`, `BATCH_SIZE`, `EPOCHS`).

2. **Train Model on Kaggle/Colab:**



   * Upload scripts and dataset to Colab.

   * Enable GPU runtime.

   * Run:



     ```bash

     python model_training.py

     ```

3. **Best Model:** Training checkpoints and the best-performing model will be saved as `best_model.h5`.



---



## Inference Pipeline



Use `inference.py` to load the trained model and classify new images:



```bash

python inference.py --image path_to_image.jpg

```



This script outputs the predicted class and confidence score. Integrate with the factory camera feed for real-time classification.



---



## Streamlit Dashboard



Launch the interactive dashboard to upload images and visualize predictions:



```bash

streamlit run dashboard.py

```



Features:



* Image uploader

* Real-time classification display

* Confidence visualization



---



## Advanced Data Processing with DeepSeek



If you have access to DeepSeek:



```bash

python deep_insights.py --csv production_data.csv

```



This generates a comprehensive report on throughput, waste metrics, and optimization suggestions.



---



## Testing and Optimization



* **Local Testing:** Validate the inference pipeline with sample images.

* **Fine-Tuning:** Unfreeze top layers of EfficientNet for additional training.

* **Quantization:** Use TensorFlow Lite for edge deployment.

* **Batch Processing:** Optimize throughput by processing images in batches or asynchronously.



---



## Deployment



* **REST API (Optional):** Wrap inference in FastAPI:



  ```bash

  uvicorn api:app --host 0.0.0.0 --port 8000

  ```

* **Docker:** Containerize the app for scalable deployment.

* **Cloud Services:** Deploy on AWS EC2, GCP Compute Engine, or Azure VM.



---



## Contributing



Contributions are welcome! Please follow these steps:



1. Fork the repository

2. Create a feature branch (`git checkout -b feature-name`)

3. Commit your changes (`git commit -m 'Add feature'`)

4. Push to the branch (`git push origin feature-name`)

5. Open a Pull Request



---



## License



This project is licensed under the MIT License. See the [LICENSE](LICENSE) file for details.



---



## Acknowledgments



* [TrashNet Dataset](https://github.com/garythung/trashnet) for waste classification data

* TensorFlow & Keras for deep learning APIs

* Streamlit for rapid dashboard development

* DeepSeek for advanced data analytics



---



Happy optimizing!