Developed by our partner AUA, the document presents a set of artificial intelligence (AI) and machine-learning models designed to support the design, optimisation, and assessment of PHA-based biopolymers. These materials are at the core of ANIPH’s strategy to reduce the environmental impacts of plastic products used in humanitarian contexts.

Designing biodegradable plastics that are both high-performing and safe is a complex process. Materials must meet functional requirements while also degrading appropriately at the end of life and posing no risk to human health or ecosystems.

The ANIPH AI predictive tool is developed to address this challenge by enabling early-stage prediction of key material properties, reducing reliance on lengthy and resource-intensive laboratory testing and supporting Safe and Sustainable by Design (SSbD) decision-making.

The deliverable describes AI models capable of predicting:

• Thermal and rheological properties, helping researchers understand how materials behave during processing and use
• Biodegradation behaviour, estimating how materials break down in different environments
• Cytotoxicity and ecotoxicity indicators, supporting early assessment of potential health and environmental risks

These predictions allow researchers to compare material formulations, identify promising candidates, and avoid designs that may lead to unwanted impacts later in development.

The results show that machine-learning approaches can successfully model complex relationships between PHA composition and material performance, biodegradability, and safety indicators. The study demonstrates that AI tools can play a reliable supporting role in guiding material development, accelerating innovation while maintaining a strong focus on sustainability and safety.

The AI predictive tool contributes directly to ANIPH’s broader goal of replacing fossil-based plastics used in humanitarian products, such as medical wound dressings and water-barrier packaging, with biobased, biodegradable solutions that perform effectively even in contexts where waste management infrastructure is limited.

In the next stages of the project, the tool will be further integrated into ANIPH’s digital and decision-support framework, strengthening traceability and transparency across the material life cycle.

Cover photo by charlesdeluvio on Unsplash