genPI · Generative Polymer Informatics

Embarking on a transformative journey, this groundbreaking research proposal unveils generative polymer informatics (genPI)-a revolutionary paradigm that leverages the unprecedented capabilities of transformer-based artificial intelligence to catapult polymer discovery, design, and development into a new era. In a time when sustainable and advanced materials are more crucial than ever, genPI aims to overcome key challenges in polymer science by exploring vast chemical spaces, accurately predicting properties, generating novel structures with targeted functionalities, and forecasting synthesis feasibility. The project is strategically organized into three interconnected work packages: rapid exploration of polymer chemical spaces (WP1), creation of customized polymer structures (WP2), and prediction of synthesis routes (WP3). These are supported by transversal activities that ensure data extraction, workflow integration, and democratization of tools and knowledge. By harnessing machine learning techniques, genPI endeavors to invert the traditional polymer design paradigm, empowering the creation of polymers with precisely engineered properties. This pioneering approach holds the promise of unveiling entirely new classes of polymers with unprecedented characteristics, directly addressing pivotal issues such as environmental sustainability, advanced functionality, and innovative synthesis methods. The project's inherently interdisciplinary nature-melding materials engineering, polymer science, and artificial intelligence-positions it at the vanguard of materials science innovation. Its successful implementation could revolutionize industries from sustainable manufacturing to healthcare, paving the way for a new era in polymer science aligned with global sustainable initiatives. This research not only promises to accelerate polymer discovery but also democratizes access to the advanced built tools, fostering collaboration and innovation across academia and industry.