University of Freiburg collaborates with Bosch Center for Artificial Intelligence

The technology company Bosch has decided to fund the basic research of Frank Hutter, Professor of Machine Learning at the Faculty of Engineering of the University of Freiburg, with 6.4 million euros until the end of 2022 as part of a new collaboration agreement. Together with his team, Hutter, who is considered to be one of the world’s founders of Automated Machine Learning (AutoML) research, won the world champion title in the international AutoML competition in 2015-16 and 2017-18. He presented the joint project with Bosch at this year’s “Robotics: Science and Systems” conference in Freiburg. Bosch uses AutoML technologies to automate the training of artificial intelligence (AI) methods at its Bosch Center for Artificial Intelligence (BCAI). By doing so, they aim to radically simplify the application of new AI approaches and thereby accelerate the transfer of basic research to applications.

“The University of Freiburg is a beacon of artificial intelligence worldwide,” said Rector Prof. Dr. Hans-Jochen Schiewer. “We’re excited that Bosch, one of the leading industrial companies in AI research, has chosen the University of Freiburg. This emphasizes the strength of both our excellent researchers and of the University of Freiburg as a center of AI research.”

“With Frank Hutter, we now have a leading global expert in automated machine learning. Our collaboration will also help to translate BCAI’s research results to the entire company,” said Dr. Christoph Peylo, Global Head of BCAI.

Hutter’s basic research project focuses on making the application of machine learning easier and more effective. Machine learning is a form of artificial intelligence in which a program autonomously searches for connections within the data it has been fed. Any kind of data can be used – for example, from meteorology, medicine, or the neurosciences. Recently, deep learning – a process in which machine learning is based on artificial neural networks – has been particularly successful. The application of machine learning, especially deep learning, requires expertise in the field, however. “Our work in AutoML is all about democratizing machine learning and ensuring that it is available for non-experts to use – for example, in small and medium-sized businesses,” Hutter stressed. The research project’s findings will continue to be published in scientific journals and be made available as open source software for anyone to use. Hutter independently designed the research program based on a shared vision for the field of AutoML research. The project will create research positions for up to twelve doctoral candidates at the University of Freiburg.

Bosch aims to harvest some of the findings of Hutter’s basic research and develop these in BCAI’s product-based research & development. Possible applications for automated machine learning include robotics, self-driving cars, sensor data analysis for predictive maintenance, as well as image and language recognition.