• WP2

AI and machine learning for PFAS prediction

WP2 connects chemical, environmental, consumer-product and industrial data to support prediction and decision-making.

LEAD

UNIWA

PROJECT MONTHS

M1–M42

  • 01

Integrate and prepare data

Harmonise PFAS information from WP1 with environmental, regulatory and industrial datasets so models can use comparable inputs.

  • 02

Predict occurrence and sources

Estimate PFAS occurrence and concentrations across water, soil, indoor dust and air, and identify likely contamination sources.

  • 03

Estimate toxicity

Use QSAR and graph-based neural networks to screen PFAS with limited experimental toxicology data and identify structural features of concern.

  • 04

Project environmental fate

Combine AI and environmental process models to simulate PFAS transport and accumulation over 5–10 years under different regulatory scenarios.