---
title: PFAS-SBEAD AI Optimization Pipeline
emoji: ⚗️
colorFrom: blue
colorTo: purple
sdk: docker
app_port: 8501
pinned: false
---

# Closed-Loop PFAS-SBEAD Optimization Pipeline

AI-driven optimization platform for PFAS degradation using Sidestream Bioelectrochemical Anaerobic Digestion (SBEAD) reactor system.

## Features

- **PFAS Degradation Prediction**: XGBoost regressor predicting degradation percentage
- **Fluoride Release Modeling**: Random Forest for fluoride release estimation
- **Short-Chain Risk Classification**: Gradient Boosting classifier for short-chain PFAS accumulation risk
- **Reactor Instability Detection**: Isolation Forest anomaly detection for stability monitoring
- **Closed-Loop Optimization**: Bayesian-inspired recommendations for next experiments
- **Mass Balance Analysis**: Complete PFAS mass balance accounting (adsorption vs degradation)
- **Sensitivity Analysis**: Feature correlation and SHAP-proxy importance rankings

## AI Objective Function

```
AI Score = 0.40 × PFAS degradation + 0.30 × fluoride release
         - 0.15 × short-chain PFAS risk - 0.10 × energy input - 0.05 × instability
```

## Dataset

120 experimental records covering:
- Reactor parameters (OLR, HRT, pH, temperature, COD, VFA, alkalinity)
- Electrochemical parameters (voltage, current, current density, conductivity)
- PFAS parameters (initial/final concentration, degradation, adsorption)
- Degradation indicators (fluoride release, defluorination, short-chain formation)
- Stability indicators (pH drop, VFA accumulation, ORP drift, current instability)