A new class of devices exploiting Fano resonances and with important applications in information technology is suggested. Typically, the resonance of a system is described by a frequency and a lifetime, leading to a Lorentzian lineshape function. If the system instead involves interference between a discrete resonance and a continuum, a Fano lineshape appears with fundamentally different characteristics. Here, the Fano resonance is used to make a novel integrated mirror, enabling realization of Fano lasers, Fano switches and quantum Fano devices. These devices challenge well-accepted paradigms for photonic devices. The goals of the project are to demonstrate a laser with modulation bandwidth greatly exceeding all existing lasers; a nanolaser with linewidth three orders of magnitude smaller than existing nanocavity lasers; and a switch that operates at femtojoule energies and provides gain. Such devices are important for realizing high-speed optical interconnects and networks between and within chips. An increasing fraction of the global energy consumption is being used for data communication, and photonics operating at very high data rates with ultra-low energy per bit has been identified as a key technology to enable a sustainable growth of capacity demands. Existing device designs, however, cannot just be scaled down to reach the goals for next-generation integrated devices. The Fano mirror will also be used to demonstrate control at the single-photon level, which will enable high-quality on-demand single-photon sources, which are much demanded devices in photonic quantum technology. These devices all rely on the unique properties of the Fano mirror, which provides a new resource for ultrafast dynamic control, noise suppression and ultra-low energy operation. Using photonic crystal technology the project will achieve its goals in a concerted effort involving development of new theory, new nanofabrication techniques and advanced experiments.
Scientific Challenge: Immunotherapy has revolutionized cancer treatment, yet only a minor fraction of patients respond to frequently used immunotherapeutic treatments. T cell recognition of peptide-major histocompatibility (pMHC) class I complexes is essential to maintain immune surveillance and eliminate cancerous cells. Numerous products of genetic and epigenetic alterations can serve as targets for T cell recognition of cancer, yet our capacity to predict what MHC-embedded targets T cells can recognize on the surface of cancer cells is still poor, with a less than 5% hit rate. While we have robust tools for prediction of antigen presentation, we still have very limited understanding of the factors driving immunogenicity – i.e. which of the presented targets will give rise to a T cell recognition. A fundamental mechanism influencing T cell recognition is molecular mimicry. It has long been proposed that the ability of a given T-cell receptor (TCR) to recognize multiple different pMHC complexes is essential to provide immunological coverage of all potential pathogens that we may encounter. T cell epitopes, that at first glance appear very different, may have structural similarities once embedded in the MHC I binding groove, and hence appear similar to the given TCR (molecular mimicry). Objective: In MIMIC, I will determine the role of molecular mimicry in T cell recognition and demonstrate how pre-existing immunity may shape the T cell recognition of cancer antigens. I will use the SARS-CoV2 infection as a model system to understand molecular mimicry, and apply the learnings from this to cancer immunogenicity. Expected outcome: I predict that by understanding the influence of molecular mimicry, the rules governing the immunogenicity of T cell epitopes can be determined and the selection of antigens optimized - this will be essential to develop precision T cell therapies targeting tumor antigens of relevance for the individual patient.
Measures against global warming require disruptive changes in the electricity sector. Drastically reducing CO2 emissions involves replacing bulk generation units with millions of renewable energy sources, along with a rapid increase of electricity demand. Maintaining the stability of the system with current approaches becomes not only computationally intractable, but also extremely costly. Recently proposed data-driven methods have been shown to handle the sheer complexity and have an impressive performance, achieving higher accuracy while being 250-1000 times faster than traditional methods. However, power systems are safety-critical systems, where data-driven methods will never be applied if they remain a black-box. This proposal removes the barriers for the application of data-driven approaches in power system problems, proposing methods that exploit the underlying physical properties of power systems. We propose the development of physics-aware verifiable neural networks and a neural network training procedure that can supply by-design guarantees of the neural network prediction accuracy. Accuracy does no longer need to be a statistical metric. Instead, our methods can supply a provable upper bound of the prediction error over the whole input space, that the power system operators can trust. We further show how neural networks can capture non-linear constraints impossible to capture before, and can reduce any non-linear optimization problem to a tractable mixed-integer linear program with verified accuracy, potentially boosting computation speed and tractability. From a power systems context, this enables us to treat power system dynamics and optimization in a unified framework that accurately captures the true feasible region, removes various approximations, and eliminates redispatching costs, saving billions of euros per year. The proposed methods naturally extend beyond power systems, finding application to a wide range of physical safety-critical systems.