Data Science Makes Chemistry Smarter

One of the most significant challenges facing the chemical industry is developing new materials and products faster and more sustainably, and integrating them into the raw materials cycle. How can digital tools help improve research and production by communicating with customers? How can data science and artificial intelligence contribute to more sustainable, resource-efficient chemistry? How can operations and supply chain management be optimized through increased data usage in conjunction with machine learning? Experts from leading established companies, start-ups and universities discussed these questions at Chemalytix in Leverkusen, the first data science conference in the chemistry field. High-performance materials producer Covestro launched this event together with Bayer and Evonik. Covestro's first international data science hackathon, previously held at three universities worldwide, was won by a student team from RWTH Aachen University with an advanced production process model based on machine learning. At the conference in Leverkusen, the students were honored and received the global first-place award.

Identifying synergies and leveraging potential

Covestro Deputy General Manager of Commerce Sucheta Govil said "We have not yet fully utilized all the potential of these new technologies. Our industry must continue to develop interaction between people and artificial intelligence, and ensure that data and intellectual property remain secure. If we are to continue succeeding in the long term, we must overcome this challenge now" and added "We can only find answers by working together. Therefore, it is very important to collaborate with customers as well as with leading partners in the sector." Kemal Malik, Member of the Innovation Management Board at Bayer, said "Data science has the potential to significantly transform our entire value chain, from research and development to production and logistics to sales and marketing. We see the tremendous potential of data science in the life sciences, for example in digital agriculture and precision medicine. At the same time, digital technologies will play a key role in helping us achieve our global sustainability goals. Partnerships with universities, start-ups and other large companies are very important in this dynamic environment." Henrik Hahn, Head of Digital at Evonik, said: "I view Data Science as a literally application-oriented science and therefore much more than a Business Intelligence Ferrari. While BI enables better decisions based on what is known, data science begins with the unknown. This can provide decisive competitive advantages in chemistry. At Evonik, therefore, we have established central contact points for data science. To put it briefly; we are looking at how we can jointly develop a strong magnet. We want to use it to find the famous needle in the haystack beforehand. Collecting the right data with the right tools – that is all there is to it."

Hackathon winners announced

Practical questions from supply chains and chemical industry production were solved by approximately 90 data and engineering science students in the first global hackathon. Covestro requested competition simultaneously for three days in November from leading universities: Carnegie Mellon University in Pittsburgh, Tongji University in Shanghai and RWTH Aachen. The winner with the best solution was the "Datavengers" team from RWTH Aachen, which received the award in Leverkusen. "Students developed new answers using different machine learning methods. Nils Janus, Head of Advanced Data Analytics at Covestro, said "They demonstrated that data and proper usage can make the entire industry smarter and faster. We are showing that the chemical industry can be an exciting employer for smart data scientists."