Reinhart Koselleck project of German Science Foundation (DFG)
Extraordinary success for the Faculty of Electrical Engineering and Information Technology, TU Dresden: Another Reinhart Koselleck Excellence Project for Professor Juergen Czarske was approved.
The Koselleck Project of the German Research Foundation (DFG) not only includes a substantial funding amount of €1,525,000, but also offers considerable scope to pursue particularly innovative research with paradigm shifts over five years. Although the new Koselleck Project is just beginning, Prof. Czarske was already honored for a Koselleck Project on a completely different topic in 2014. Two Koselleck Project Awards have only been presented twice in the history of the DFG.
The low-bureaucracy funding with global funds, which can be used flexibly for personnel, material, or capital expenditures, allows for extremely high freedom for innovative basic research. Reinhart Koselleck projects require not only a high-risk topic in the positive sense, but also an outstanding scientific CV. Prof. Juergen Czarske is a world-leading researcher in electro-optical systems engineering who has received prestigious international awards for his groundbreaking achievements, including the Joseph Fraunhofer Award, in Washington D.C. (OPTICA, Society for Advancing Optics and Photonics Worldwide, USA), the IEEE Laser Instrumentation Award, the Chandra S. Vikram Award in Optical Metrology, and the Dennis Gabor Award in Holography, both presented in San Diego by SPIE (International Society of Optics and Photonics, USA).
The Reinhart Koselleck project "Physics-Informed Deep Learning Systems for Secure Information Transmission with Multimode Fibers" is not about usual suppressing light scattering for distortion-free information transmission with single-mode fiber, but rather about using it in a targeted manner to exploit scattering in multimode-fibers. This new approach encompasses artificial intelligence (AI), physical modeling, computational imaging, neuromorphic computing, metamaterials/nanotechnology, and second-generation quantum technology. Physics-informed deep learning systems are increasingly proving to be a universally applicable Swiss Army knife for applications such as and optical communication and sensing, automated driving, intelligent (endo)-microscopy for clinics and biomedicine, virtual reality (VR), augmented reality (AR) / metaverse, advanced manufacturing / production technology, quantum communication, quantum imaging, quantum computing.
With this project, Prof. Czarske, together with an interdisciplinary team, aims to introduce a radically new paradigm for scattering media such as multimode fibers: the use of AI to capture light scattering, which not only corrects distortions but also enables the scattering itself to be used as a security code. The risk of the project lies in precise measurements of complex linear and nonlinear scattering processes in fibers, which can also change rapidly. However, if this challenge is met, fundamentally novel applications will arise not only for classical and quantum communication, but also for biomedicine, such as fiber endoscopes for understanding the physics of life. However, there is considerable uncertainty as to whether this transfer from AI methods to quantum methods is even possible, which is why this research topic can only be investigated within the framework of the Koselleck project and not with conventional project funding. Furthermore, an interdisciplinary team of experts from various disciplines, such as systems engineering, electrical engineering, physics, quantum technology, computer science, and biomedicine, working closely together over a long period of time, is required. While there is no guarantee of project success, with new approaches, a commitment to independent research over five years, as well as diligence, perseverance, and an excellent team, paradigm shifts can be achieved, and sensational breakthroughs at Terra incognita are possible. An exciting time has started.