Try out the application which is live
A machine learning-powered web application that helps optimize delivery routes and supply chain decisions based on carbon emissions. This project predicts the CO₂ equivalent per USD spent using emission factor data and provides businesses a way to move toward eco-friendly, sustainable logistics.
Carbon Footprint Optimizer for Supply Chain Logistics, smart deep learning-powered tool to predict carbon emissions in supply chains based on emission factors and margin data.
Traditional logistics systems focus on cost and time, often ignoring environmental impact. This tool aims to change that by using data-driven predictions to estimate carbon emissions associated with different segments of the supply chain.
By entering key emission factor details, users can view estimated emissions and make better decisions for sustainability.
- ✅ Predict carbon emissions per dollar spent using a trained ML model
- ✅ Interactive UI built with HTML + Flask
- ✅ Clean evaluation metrics: MAE, RMSE, % error
- ✅ Live deployment (via Netlify)
- ✅ Supports green decision-making in logistics and supply chain
| Layer | Technology |
|---|---|
| Backend | Python, Flask |
| ML Model | Scikit-learn, RandomForestRegressor |
| Frontend | HTML, CSS (basic) |
| Deployment | Netlify (frontend), Flask (local backend) |
- Model: Random Forest Regressor
- Inputs:
- Supply Chain Emission Factors (without margins)
- Margins of Emission Factors
- Target:
- Total Supply Chain Emission Factors (with margins)
- Metrics:
- MAE: ~0.04
- RMSE: ~0.51
- Avg % Error: ~1.23%
Special thanks to the medical and AI communities for their valuable datasets and research.
Inspirational guidance from Dr. Victor Ikechukwu. Explore their work: Dr. Victor Ikechukwu.
This project is licensed under the MIT License.
🔥 If you like this project, don't forget to ⭐ it on GitHub!