ERC STG - genPI

The ERC funded project generative polymer informatics (genPI) uses transformer-based artificial intelligence to accelerate the discovery and design of polymers. As the demand for sustainable and advanced materials grows, genPI aims to solve major challenges in polymer science. It does this by exploring vast chemical polymer spaces, accurately predicting material properties, designing novel structures with specific functions, and determining how to synthesize them.

The project focuses on four main goals: rapidly exploring vast chemical polymer spaces, accurately predicting material properties, creating customized polymer structures, and predicting synthesis routes. These efforts are supported by a framework that ensures data is easily extracted, workflows are integrated, and tools are accessible to everyone. By using machine learning, genPI also reverses the traditional design process, allowing researchers to create polymers with precise, pre-engineered properties. This interdisciplinary mix of materials engineering, polymer science, and AI aims to uncover new classes of materials that support environmental sustainability and healthcare, while making advanced research tools available to both academia and industry.