Created by Team CodeBlooded during 418 hackathon hosted by Enigma under AEON 2025 Bridging the gap between πΎ farmers, π sugar mills, and π distributors using AI-powered sugarcane quality prediction and real-time trading.
π Live Website: https://codeblooded-xi.vercel.app/
In the sugar industry, farmers are often underpaid due to lack of quality measurement and transparency. SugarMommy solves this by
- π€ Predicting sugar content (TS%) using NIR spectroscopy and machine learning
- π Connecting farmers, mills, and distributors under one seamless platform
- πΈ Enabling fair, dynamic pricing based on quality β no middlemen, no friction
- π¦ Providing inventory and order management
- Upload sugarcane data via CSV or web form
- Automated prediction of Total Sugar (TS%)
- Receive a quality grade (AβE) and price recommendation
- List sugarcane batches on the marketplace in real time
- Browse listings by:
- Quality Grade (AβE)
- Predicted TS%
- Price per ton
- Buy directly from farmers
- View quantity and location before purchase
- Model: Partial Least Squares Regression (PLS)
- Features: 232 NIR wavelength amplitudes (900β1700 nm)
- Target: Total Sugar (TS%)
- Pre-processing: Standardization + IQR outlier removal
- Accuracy: ~95% (RΒ² Score)
| Grade | TS% Range |
|---|---|
| A | Above 41.69 |
| B | 34.24 β 41.69 |
| C | 17.4 β 34.24 |
| D | 4.76 β 17.4 |
| E | Below 4.76 |
π Live Model: Try it on Hugging Face
π Live Website: https://codeblooded-xi.vercel.app/
| Layer | Tech Used |
|---|---|
| Frontend | Next.js + TypeScript |
| Backend | Firebase |
| ML Inference | Scikit-learn (PLS Regression) |
| ML UI | Gradio |
| Database | Firestore |
| Hosting | Vercel |
Clone the repo and install dependencies:
git clone https://github.com/your-username/sugarmommy.git
cd sugarmommy
npm install
npm run dev