A scientist from Ulm made a big splash in the “Science Year 2019”, which focuses on artificial intelligence. Professor Susanne Biundo-Stephan, director of the Institute of Artificial Intelligence, was named one of the “ten influential minds” in the history of German artificial intelligence. She was selected by an expert panel from the German Informatics Society within the scope of the Science Year, put on by the Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung: BMBF).
According to the jury, Biundo-Stephan is one of the central minds of German artificial intelligence research in the area of “automated planning”, with a focus on the areas of “hierarchical planning” and “cognitive technical systems”. Biundo-Stephan, who is originally from Rhineland-Palatinate, has also conducted research on the topics of “knowledge modelling” and “automated reasoning”. What is it about the field of AI that intrigues the local researcher so much? “I find it fascinating that human beings are capable of creating technical systems that behave intelligently and exhibit cognitive abilities”, she reveals.
Professor Biundo-Stephan doesn’t believe that “strong AI” - artificial intelligence with its own conscious mind and acting on its own volition - will be emerging any time in the near future. She’s more than sceptical about the Human Brain Project, i.e. replicating the human brain by technical means. “I think it’s pretty unrealistic and also rather arrogant. However, in certain areas, cognitive technical systems are without a doubt far superior to us, and are certainly capable of assisting humans”, explains the computer scientist who researches at Ulm University with a focus on human-technology interaction. One example of such supporting technologies is so-called companion systems, which support humans in operating technical devices. These are oriented toward humans and their individual abilities as well as their preferences and needs. In an ideal scenario, such systems can even adapt to the current environment and emotional state of their users.
Adapted to the user, the system reveals differing solutions
What appears to be such an obvious idea is actually quite remarkable from a technical perspective. Automated action planning requires the defining of initial conditions, goals, individual steps and quality criteria in order for a recommendation for a course of action to be created, but that’s not all. The planning process must also include causalities, as well as temporal and hierarchical relationships between the individual substeps. The system is adapted to fit the individual use and then creates different solutions or results. “Among other things, we draw on formal planning modules and conclusion processes, which can help to explain how certain recommendations for action arise and why a certain result makes more sense than another one under certain conditions”, relates Biundo-Stephan, spokesperson for the now discontinued Collaborative Research Centre on companion technologies. The Institute is currently running a transfer project funded by the German Research Foundation (Deutsche Forschungsgemeinschaft: DFG), in cooperation with the company Bosch, which is using technologies developed in the CRC to provide intelligent support for craftsmen.
In terms of the international significance of German AI research, the scientist who completed her PhD in the field of “automated reasoning” is confident: “Germany has always been at the top in the area of knowledge- and logic-based processes. That hasn’t changed in the last few years”, says the director of the Institute of Artificial Intelligence. The frequently mentioned competition from the USA and China refers more to applications in the area of pattern recognition processes in machine learning, such as “deep learning”.
Susanne Biundo-Stephan studied computer science in Kaiserslautern and Karlsruhe and received her doctorate from Karlsruhe University. Following her studies, she led a research team at the German Artificial Intelligence Research Centre (Deutsches Forschungszentrum für Künstliche Intelligenz: DFKI) in Saarbrücken for nearly ten years. In 1998, she was the first woman to be offered a professorship in computer science at Ulm University. Biundo-Stephan was the initiator and coordinator of the “European Network of Excellence in AI Planning” (PLANET). As the spokesperson for the Transregional Collaborative Research Centre 62 “Companion Technology for Cognitive Technical Systems”, which she headed at Ulm University from 2009 to 2017, she was the first woman in Germany at the head of a computer science CRC. In addition, she is a founding member of the most prestigious international conference in the field of intelligent action planning and a fellow of the European Association for Artificial Intelligence (EurAI). Since the end of 2017, she has been the gender diversity officer at Ulm University, promoting having more women in academics and on university committees. Of the “ten influential minds” in AI history, two are female.
The German Informatics Society participated in the Federal Ministry of Education and Research (BMBF) “Science Year 2019” with an anniversary project on the history of artificial intelligence. Within the scope of the project “#KI50: Artificial Intelligence in Germany – Yesterday, Today, Tomorrow”, “ten significant technologies” in artificial intelligence history were selected, along with the “ten influential minds” (https://ki50.de/).
Text and Mediacontact: Andrea Weber-Tuckermann