ERC FUNDED PROJECTS

"CON-HUMO focuses on novel concepts for automatic control based on data-driven human models and machine learning. This enables innovative control applications that are difficult if not impossible to realize using traditional control and identification methods, in particular in the challenging area of smart human-machine interaction. In order to achieve intuitive and efficient goal-oriented interaction, anticipation is key. For control selection based on prediction a dynamic model of the human interaction behavior is required, which, however, is difficult to obtain from first principles. In order to cope with the high complexity of human behavior with unknown inputs and only sparsely available training data we propose to use machine-learning techniques for statistical modeling of the dynamics. In this new field of human interaction modeling – data-driven and machine-learned – control methods with guaranteed properties do not exist. CON-HUMO addresses this niche. Key methodological innovation and breakthrough is the merger of probabilistic learning with model-based control concepts through model confidence and prediction uncertainty. For the sake of concreteness and evaluation the focus is on one of the most challenging problem classes, namely physical human-machine interaction: Because of the physical contact between the human and the machine not only information, but also energy is exchanged posing fundamental challenges for real-time human-adaptive and safe decision making/control and requiring provable stability and performance guarantees. The developed methods are a direct enabler for societally important applications such as machine-based physical rehabilitation, mobility and manipulation aids for elderly, and collaborative human-machine production systems. With its fundamental results CON-HUMO lays the ground for the systematic control design for smart human-machine/infrastructure interaction."

1 494 640 €

Start date: 2014-02-01, End date: 2019-01-31

This research investigates a new measure that arises in information theory called directed information. Recent advances, including our preliminary results, shows that directed information arises in communication as the maximum rate that can be transmitted reliably in channels with feedback. The directed information is multi-letter expression and therefore very hard to optimize or compute. Our plan is first of all to find an efficient methodology for optimizing the measure using the dynamic programming framework and convex optimization tools. As an important by-product of finding the fundamental limits is finding coding schemes that achieves the limits. Second, we plan to find new roles for directed information in communication, especially in networks with bi-directional communication and in data compression with causal conditions. Third, encouraged by a preliminary work on interpretation of directed information in economics and estimation theory, we plan to show that directed information has interpretation in additional fields such as statistical physics. We plan to show that there is duality relation between different fields with causal constraints. Due to the duality insights and breakthroughs in one problem will lead to new insights in other problems. Finally, we will apply directed information as a statistical inference of causal dependence. We will show how to estimate and use the directed information estimator to measure causal inference between two or more process. In particular, one of the questions we plan to answer is the influence of industrial activities (e.g., $\text{CO}_2$ volumes) on the global warming. Our main focus will be to develop a deeper understanding of the mathematical properties of directed information, a process that is instrumental to each problem. Due to their theoretical proximity and their interdisciplinary nature, progress in one problem will lead to new insights in other problems. A common set of mathematical tools developed in

1 224 600 €

Start date: 2013-08-01, End date: 2019-07-31

Optical nanoscopy has given a glimpse of the impact it may have on medical care in the future. Slow imaging speed and the complexity of the current nanoscope limits its use for living cells. The imaging speed is limited by the bulk optics that is used in present nanoscopy. In this project, I propose a paradigm-shift in the field of advanced microscopy by developing optical nanoscopy based on a photonic integrated circuit. The project will take advantage of nanotechnology to fabricate an advance waveguide-chip, while fast telecom optical devices will provide switching of light to the chip, enhancing the speed of imaging. This unconventional route will change the field of optical microscopy, as a simple chip-based system can be added to a normal microscope. In this project, I will build a waveguide-based structured-illumination microscope (W-SIM) to acquire fast images (25 Hz or better) from a living cell with an optical resolution of 50-100 nm. I will use W-SIM to discover the dynamics (opening and closing) of fenestrations (100 nm) present in the membrane of a living liver sinusoidal scavenger endothelial cell. It is believed among the Hepatology community that these fenestrations open and close dynamically, however there is no scientific evidence to support this hypothesis because of the lack of suitable tools. The successful imaging of fenestration kinetics in a live cell during this project will provide new fundamental knowledge and benefit human health with improved diagnoses and drug discovery for liver. Chip-based nanoscopy is a new research field, inherently making this a high-risk project, but the possible gains are also high. The W-SIM will be the first of its kind, which may open a new era of simple, integrated nanoscopy. The proposed multiple-disciplinary project requires a near-unique expertise in the field of laser physics, integrated optics, advanced microscopy and cell-biology that I have acquired at leading research centers on three continents.

1 490 976 €

Start date: 2014-02-01, End date: 2019-01-31

With this ERC project I want to induce a paradigm shift in the development of integrated nonlinear optical devices. Nonlinear optics, the scientific discipline in which nonlinear light-matter interactions are studied, has been a very active area of research ever since the invention of the laser in 1960. Although this scientific branch has great application potential when implemented in on-chip optical waveguides, its promise for the development of widely usable integrated optical devices has not yet been fulfilled. The state-of-the-art of integrated nonlinear optical devices indeed does not comply with the requirements for widespread deployment as these devices rely on non-standard waveguide designs, large on-chip foot prints and/or impractical pump lasers. Therefore, I propose in this project to eliminate the issues of the state-of-the-art devices by introducing novel material and device physics. More specifically, my goal is to exploit extreme, but practically unexplored, nonlinear optical properties of graphene-covered silicon waveguides to develop next-generation near-infrared-pumped nonlinear supercontinuum light sources. These will truly be “next-generation” sources as they will rely on standard waveguide design, ultra-compact foot prints and practical near-infrared pump lasers, while exhibiting unprecedented performances. The concrete objectives of my project are to theoretically study, model, fabricate and experimentally demonstrate three novel graphene-on-silicon-based nonlinear optical devices that rely on three different nonlinear optical effects, and the on-chip cascading of these novel devices to create the targeted “next-generation” near-infrared-pumped supercontinuum sources with up to four emission bands. Based on my theoretical and experimental research experience with nonlinearities in waveguides and my preliminary modeling results supporting the feasibility of these objectives, I believe that, with this ERC starting grant, I will be able to carry out this original “high-gain/high-risk” project. By doing so, I will introduce a paradigm shift in the development of integrated nonlinear optical devices enabling them to fulfill their long-awaited promise, and at the same time initiate a new era in the research on graphene and its nonlinear optical applications.

1 477 980 €

Start date: 2013-10-01, End date: 2018-09-30