ARTIFICIAL INTELLIGENCE
Leveraging AI to power sustainable innovation
Daphne Vlastari, Head of Communications and Government Relations, Europe North, BASF
Artificial intelligence (AI) is no longer an experimental addon for industry — at BASF, one of the world’s largest chemical companies, it is a strategic accelerator for creating value at scale. AI is being deployed across the company to boost productivity, accelerate innovation and improve safety, while creating new pathways to more sustainable chemistry.
For more than a decade, BASF has invested in machine learning and AI capabilities, building digital hubs worldwide that combine domain experts with data scientists, cloud and software engineers, and specialists in image analysis, forecasting and natural language processing. This cross-disciplinary capability allows BASF to pilot fast, scale what works, and embed AI into everyday business functions — from production and procurement to sales, marketing and, crucially, research and development.
Bringing AI into R&D In R&D, AI and machine learning are now considered a core part of the scientist’s toolkit: they can be applied in the prediction of new molecules and materials and their properties, in the evaluation of image data and texts, or in the automation of research laboratories.
Machine learning models accelerate the identification of promising molecules and materials by predicting properties and prioritising candidates for laboratory testing. Image and text analytics speed the extraction of insights from microscopy, spectroscopy and scientific literature. Combined with laboratory automation and high throughput experimentation, these tools dramatically shorten iteration cycles
20
and increase the probability of success for new formulations and catalysts.
At BASF we aim to harness and translate the latest AI developments into tangible solutions that optimise our value chains and accelerate innovation. Our approach is pragmatic: we focus first on clear value cases, confirm benefits through rapid pilots, and then scale successful solutions across divisions. For example, generative AI and knowledge retrieval systems such as QKnows AI are already helping R&D teams find internal reports and external literature more efficiently, reducing duplication and speeding decision making.
Recognising that fundamental advances often come from collaboration, BASF is also strongly engaged in collaboration with various academic institutions in the field of AI such as the Technical University of Berlin, MIT, KU Leuven, Imperial College London, McMaster University and the University of Waterloo.
For example, through our partnership with Imperial College London, we have jointly developed a new algorithm that could boost chemical R&D. Setting up a new production line requires experiments to find the best processing conditions (temperatures, reactants, etc.), but these tests are slow and costly, so there is a need to find optimal settings in as few experiments as possible. Bayesian optimisation is a machine learning method that recommends which conditions to try next by balancing improvements near known good settings with riskier, uncertain tests that might yield better results. However, traditional Bayesian optimisation doesn’t fit chemical R&D practices—chemists often
LUBE MAGAZINE AR TIFICIAL INTELLIGENCE DECEMBER 2025
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28
