FWF funding: New large-scale research projects

More safety with AI, better sources for terahertz radiation, an atomic view of new materials and mathematics for MRI devices: the new projects at TU Wien.

Four major research projects from very different disciplines are now being made possible at TU Wien thanks to funding from the Austrian Science Fund FWF. Two new doctoral programmes are being launched - one on the topic of automated reasoning and one on future terahertz communication. The special research area TACO, in which new materials are researched, particularly with regard to future-oriented energy technologies, is also being extended. TU Wien is also involved in a specialised mathematics research area coordinated by the University of Graz.

Using logic to improve safety in AI

Electronic systems and, increasingly, artificial intelligence are with us every day - with enormous social implications. How can we ensure that self-driving cars or AI-supported decisions in justice and medicine are reliable, error-free and fair? The doctoral programme "Automated Reasoning" aims to ensure the safety of software and deep learning with the help of logic, machine reasoning and automatic analyses. It offers young researchers exciting research topics at the interface of IT security and artificial intelligence, international internships in leading research laboratories, and intensive supervision and collaboration with excellent scientists. 13 new doctoral positions are now being awarded at TU Wien as part of this programme, 10 of which are funded by the FWF. The programme is headed by Prof. Georg Weissenbacher (Institute of Logic and Computation).

Doctoral programme "FutureTHz"

Terahertz (THz) technology offers great potential for numerous applications in the fields of medical technology, security technology, sensor technology, radio communication and earth observation. However, a major obstacle to its practical use is that there are hardly any cost-effective and compact radiation sources with sufficiently high power that can also be precisely controlled in terms of phase and frequency. The doctoral programme "FutureTHz" is researching a new type of radiation source that will make precisely this possible - so-called resonant tunnelling diode oscillators.

Under the leadership of Prof Michael Feiginov (Institute of Electrodynamics, Microwave and Circuit Engineering), the necessary fundamentals are being systematically investigated. Models are being developed and tested in experiments. The FWF-funded doc.funds.connect project is creating 5 doctoral positions and is being carried out jointly with the Institute of Telecommunications and the FH OÖ (Hagenberg Campus).

New materials for our energy future

Four years ago, the research project "TACO" began using machine learning for materials research - a method for which the Nobel Prizes in Physics and Chemistry were awarded this year. Within TACO, researchers want to resolve a series of pressing questions in materials science, particularly in catalysis on material surfaces. Different size scales are combined: To understand phenomena, one often needs to go to the atomic scale, but for technological impact, one needs to apply them on a larger scale. This line of research discipline is particularly important for new energy technologies: if we want to generate or store electrical energy, for example, we need special, high-performance materials. The collaborative Ulrike Diebold (Institute of Applied Physics, TU Wien) includes researchers from the University of Vienna and has now been extended with new funding from the FWF.

Better images of the inside of the body

Many modern measurement methods only work well if you have suitable, highly complex mathematical models with which you can process and analyse your data. One example of this is magnetic resonance imaging (MRI). Magnetic signals are used to examine the body and an image is calculated on the computer from the data received. Optimised mathematical methods are required both for the appropriate generation of the signals and for the calculation of the image data - for example, to eliminate disturbing effects caused by the patient moving during the scan. The University of Graz is therefore now leading the Special Research Centre (SFB) "Mathematics of Reconstruction in Dynamical and Active Models", and TU Wien is involved in Elisa Davoli (Institute of Analysis and Scientific Computing) and her research group play an important role in the project.