Lanxess Uses AI to Produce Prepolymers

Lanxess has completed the initial phase of a project to accelerate development of custom-made formulations. In the early phase, it has been able to manufacture

prepolymers with the help of artificial intelligence (AI), toward the goal of developing tailor-made polyurethane systems with shorter lead times, even in entirely new applications with different requirements.

The Urethane Systems business unit of the Cologne, Germany-based specialty chemicals producer said it is building on knowledge gained through its partnership with materials AI company Citrine Informatics, which is regarded as an industry leader in the use of data and artificial intelligence to accelerate the development of materials and chemicals.

Before beginning the first AI-assisted production trials, Lanxess’ process experts leveraged the Citrine platform to enlarge the company’s database of prepolymer-based formulations, then linked existing empirical measurement data using a chemical process-adapted algorithm to calculate additional measurement values. As a result, only a few physical measurements were required to verify the figures determined with AI.

In a subsequent step, Lanxess said it will seek to determine how reliably the ideal formulation needed to meet customer-specific requirements can be predicted with the aid of AI. With the broader knowledge base, the amount of testing required when developing new products could potentially be thus significantly reduced, it believes.

Jörg Hellwig, head of the company’s Digitalization Initiative launched in 2017, said the partnership between Citrine and its Urethane Systems business unit demonstrates the growing role of digitalization in product development, which is increasingly becoming standard procedure at Lanxess.

The initiative’s core objectives are to develop digital business models, introduce new technologies along the value chain, develop and utilize big data and foster digital expertise among employees.