This article was published in our newsletter. Sign up here.
Hardly anything could be simpler for us humans: we stand in front of our wardrobe and pick a t-shirt out of the middle compartment. Then we head for the kitchen where we take two eggs out of the fridge. In the office later we grab a folder off the shelf. What are for us very ordinary and everyday actions pose a big challenge for robots. Although they can already pick up different things, this only works if the objects in question are very similar in shape. This is partly because robots are still unable to understand well enough what they see. They cannot really tell the difference between a t-shirt, an egg and a folder, nor can they adapt their actions to these specific items. Grabbing a t-shirt obviously involves a different type of grip than picking up an easily-breakable egg.
How can robots be made more intelligent?
But how can computers, and robots in particular, be taught to better interpret images and objects? And how can they adapt their behaviour accordingly? These questions are the main focus of the research being pursued by Dr Anna Belardinelli. A computer scientist, she works as a research associate at the Faculty of Science at the University of Tübingen. 37 years old and originally from Italy, Dr Belardinelli is a member of the Cognitive Modeling research group. “Our research work involves numerous different disciplines, so we computer scientists collaborate with other experts such as psychologists, neuroscientists, biologists and philosophers”, explains Anna Belardinelli.
Input from many scientific disciplines
Dr Belardinelli took her degree and obtained her PhD in Rome. “Even back then I realised that computer scientists won’t be able to work out how to optimise robots on their own.” She explains that Germany is very well positioned, especially when it comes to interdisciplinary work in this area, with interdisciplinary bachelor’s and master’s degree courses already on offer in Osnabrück and Tübingen, for example.
It was precisely this interdisciplinary approach that brought Anna to Germany around seven years ago. She started off in the city of Bielefeld in North Rhine-Westphalia, home since 2007 to the Cluster of Excellence Cognitive Interaction Technology (CITEC). Researchers there are working on the basic scientific principles of how to make machines more intelligent and helpful. The goal is to enable them to interact naturally with humans and adapt to changing situations.
Short journey times, international environment
When a new degree course in Cognitive Science was established some time later at the University of Tübingen, Anna Belardinelli moved to this university in the southwest of Germany. “I really like it here a lot. The city is not too big, journey times are short, and everything here is very calm and peaceful.” Nonetheless, she goes on to say that Tübingen offers a very international environment thanks to its university, and for example the Max Planck Institute for Biological Cybernetics and the Max Planck Institute for Intelligent Systems. Anna Belardinelli is relatively independent in her scientific research. “Obviously I discuss the precise topics of my research and my experiments with my boss, but I can work fairly freely when it comes to my actual research. That is something I value a great deal”, says the computer scientist.
Different career paths for young researchers
Young researchers can also find good conditions for studying computer science at graduate schools. Through its graduate schools, the German Research Foundation (DFG) is currently supporting 45 excellent and internationally competitive institutions in Germany, all of them academic leaders in their fields. Other institutions offering special PhD programmes also provide young researchers with optimal support and supervision throughout their research work. The following is a selection of graduate schools and other institutions:
- Aachen Institute for Advanced Study in Computational Engineering Science
- Bavarian Graduate School of Computational Engineering at Friedrich-Alexander-Universität Erlangen-Nürnberg and the Technical University of Munich
- Center for Doctoral Studies in Informatics and its Applications in Munich
- Fraunhofer Institute for Applied Information Technology FIT in Sankt Augustin
- Graduate School of Excellence Computational Engineering in Darmstadt
- Max Planck Institute for Informatics in Saarbrücken
- Saarbrücken Graduate School of Computer Science
According to studies like the CHE Ranking, examples of institutions well-known for their excellent support and research in computer studies also include
- Hasso Plattner Institute Potsdam
- Jacobs University Bremen
- University of Augsburg
- University of Bayreuth
- University of Magdeburg
Young computer scientists can also gain an insight into applied research in German companies. Computer science graduates can work in many different areas – for example in data processing, plant engineering and construction or in the energy sector.
Important social relevance
Anna Belardinelli talks with great passion about her work – and about what research in this area is achieving. In the future, innovative robots could provide support in the everyday routines of ill, old and mobility-impaired people, helping them to pursue active and self-determined lives. Robots could also make it easier for people to reconcile the demands of family and work life.
Robots and other cognitive systems such as cars and smartphones could simplify our lives in many areas. Who knows, perhaps one day they could even write an academic thesis. This won’t help Anna Belardinelli – she has already nearly finished her postdoctoral dissertation, which she plans to submit in 2017.