AI and computer simulations have been effectively utilized in the rubber industry to streamline product design processes. Both technologies eliminate the need for physical prototyping by predicting product performance based on design proposals, evaluating the feasibility of manufacturing designs, and enabling pre-production process planning.[br] While sharing similar objectives, AI and equation-based computer simulations possess distinct characteristics. Understanding these nuances and leveraging their strengths strategically will be crucial moving forward. This paper delves into the specific features of each technology, illustrated through relevant examples, and discusses their future prospects in the rubber industry.

The most important resource of data-driven research is data. However, the amount of data available on polymeric materials is overwhelmingly small compared to other material systems. To address this data scarcity, we describe a simulation-to-real (Sim2Real) transfer learning methodology, in which a large dataset of polymer property data generated through molecular simulations is leveraged to enhance a smaller set of experimental data. Additionally, we introduce the development of RadonPy, a Python library that fully automates all-atom classical molecular dynamics simulations of polymer properties, serving as the materials informatics (MI) foundation for Sim2Real transfer learning. We also discuss the industry-academia collaboration to develop the world's largest database of computational polymer properties using the RadonPy. As of July 2024, more than 80,000 polymers have been calculated, establishing this as the world's largest such database. Using the RadonPy database as training data, we have constructed a calibration model that compensates for the gap between calculated and experimental data through Sim2Real transfer learning. Furthermore, we observed a scaling law in Sim2Real transfer learning, where the generalization performance to experimental values improves according to a power law as the amount of simulation data increases.

In the research development of polymer science based on AI or machine learning approach, the treatment of higher-order structure of polymers is essential, but often troublesome. The complex structure or high dimensional data should be reduced to low dimensional physical/feature values, which afford interpretability for human. This review describes higher-order structure descriptor of polymers for AI research development and its application for describing rubber elasticity based on data science.

This article explains the leveraging AI named HAICoLab to open up new horizons through metacognition and creativity. It has also been shown that the purpose of using data in material and product development is feature exploration. A feature is a combination of design factors that contribute to the desired performance balance, and an example of feature discovery using clustering and machine learning with ChatGPT is presented. In addition, a tire development process based on a "human-AI collaboration" using AI and XAI (eXpalainable AI) to present features to humans is introduced. On the other hand, it is also stated that in order to utilize data and AI to gain new insights, it is preferable to reduce the influence of cognitive biases such as preconceptions. An example is introduced showing that facilitation to maintain an open mind is effective for this purpose.

Digital transformation (DX) is permeating materials design through advanced digital technologies such as databases, theoretical simulation, and artificial intelligence (AI). Materials Informatics (MI) has emerged as a pivotal approach in materials development, utilizing data to predict the properties of unknown material structures and optimize structures to achieve desired functions and properties. At Asahi Kasei, Digital Value Co-Creation integrates digital technologies across R&D, manufacturing, and business. The Informatics Initiative focuses on driving R&D DX and strengthening informatics capabilities. Its efforts include training experimental researchers in the use of MI and providing an in-house cloud-based environment (IFX-Hub) to facilitate MI implementation. Successful MI applications in polymer development demonstrate the potential of this approach, although challenges remain in handling complex data structures and ensuring data quality. Future directions include implementation of high-level MI techniques and expansion of the “smart lab” concept for autonomous experimentation and data-driven optimization. Furthermore, generative AI is a promising technology for R&D. Guidelines have been released for general users to ensure safe and secure use. System platforms are also being developed for advanced use and secure use of internal documents.