ERC FUNDED PROJECTS

Plants typically release large quantities of volatiles in response to attack by herbivores or pathogens. I may claim to have contributed to various breakthroughs in this research field, including the discovery that the volatile blends induced by different attackers are astonishingly specific, resulting in characteristic, readily distinguishable odour blends. Using maize as our model plant, I wish to take several leaps forward in our understanding of this signal specificity and use this knowledge to develop sensors for the real-time detection of crop pests and diseases. For this, three interconnected work-packages will aim to: • Develop chemical analytical techniques and statistical models to decipher the odorous vocabulary of plants, and to create a complete inventory of “odour-prints” for a wide range of herbivore-plant and pathogen-plant combinations, including simultaneous infestations. • Develop and optimize nano-mechanical sensors for the detection of specific plant volatile mixtures. For this, we will initially adapt a prototype sensor that has been successfully developed for the detection of cancer-related volatiles in human breath. • Genetically manipulate maize plants to release a unique blend of root-produced volatiles upon herbivory. For this, we will engineer gene cassettes that combine recently identified P450 (CYP) genes from poplar with inducible, root-specific promoters from maize. This will result in maize plants that, in response to pest attack, release easy-to-detect aldoximes and nitriles from their roots. In short, by investigating and manipulating the specificity of inducible odour blends we will generate the necessary knowhow to develop a novel odour-detection device. The envisioned sensor technology will permit real-time monitoring of the pests and enable farmers to apply crop protection treatments at the right time and in the right place.

2 498 086 €

Start date: 2018-09-01, End date: 2023-08-31

Recurrent neural networks (RNNs) are general parallel-sequential computers. Some learn their programs or weights. Our supervised Long Short-Term Memory (LSTM) RNNs were the first to win pattern recognition contests, and recently enabled best known results in speech and handwriting recognition, machine translation, etc. They are now available to billions of users through the world's most valuable public companies including Google and Apple. Nevertheless, in lots of real-world tasks RNNs do not yet live up to their full potential. Although universal in theory, in practice they fail to learn important types of algorithms. This ERC project will go far beyond today's best RNNs through novel RNN-like systems that address some of the biggest open RNN problems and hottest RNN research topics: (1) How can RNNs learn to control (through internal spotlights of attention) separate large short-memory structures such as sub-networks with fast weights, to improve performance on many natural short-term memory-intensive tasks which are currently hard to learn by RNNs, such as answering detailed questions on recently observed videos? (2) How can such RNN-like systems metalearn entire learning algorithms that outperform the original learning algorithms? (3) How to achieve efficient transfer learning from one RNN-learned set of problem-solving programs to new RNN programs solving new tasks? In other words, how can one RNN-like system actively learn to exploit algorithmic information contained in the programs running on another? We will test our systems existing benchmarks, and create new, more challenging multi-task benchmarks. This will be supported by a rather cheap, GPU-based mini-brain for implementing large RNNs.

2 500 000 €

Start date: 2017-10-01, End date: 2022-09-30

"Recent technological evolutions, including the cloud, the multicore, the social and the mobiles ones, are turning computing ubiquitously distributed. Yet, building high-assurance distributed programs is notoriously challenging. One of the main reasons is that these systems usually seek to achieve several goals at the same time. In short, they need to be efficient, responding effectively in various average-case conditions, as well as reliable, behaving correctly in severe, worst-case conditions. As a consequence, they typically intermingle different strategies: each to cope with some specific condition, e.g., with or without node failures, message losses, time-outs, contention, cache misses, over-sizing, malicious attacks, etc. The resulting programs end up hard to design, prove, verify, implement, test and debug. Not surprisingly, there are anecdotal evidences of the fragility of the most celebrated distributed systems. The goal of this project is to contribute to building high-assurance distributed programs by introducing a new dimension for separating and isolating their concerns, as well as a new scheme for composing and reusing them in a modular manner. In short, the project will explore the inherent power and limitations of a novel paradigm, Adversary-Oriented Computing (AOC). Sub-programs, each implementing a specific strategy to cope with a given adversary, modelling a specific working condition, are designed, proved, verified, implemented, tested and debugged independently. They are then composed, possibly dynamically, as black-boxes within the same global program. The AOC project is ambitious and it seeks to fundamentally revisit the way distributed algorithms are designed and distributed systems are implemented. The gain expected in comparison with today's approaches is substantial, and I believe it will be proportional to the degree of difficulty of the distributed problem at hand."

2 147 012 €

Start date: 2014-06-01, End date: 2019-05-31

The attoclock is a powerful, new, and unconventional tool to study fundamental attosecond dynamics on an atomic scale. We established its potential by using the first attoclock to measure the tunneling delay time in laser-induced ionization of helium and argon atoms, with surprising results. Building on these first proof-of-principle measurements, I propose to amplify and expand this tool concept to explore the following key questions: How fast can light liberate electrons from a single atom, a single molecule, or a solid-state system? Related are more questions: How fast can an electron tunnel through a potential barrier? How fast is a multi-photon absorption process? How fast is single-photon photoemission? Many of these questions will undoubtedly spark more questions – revealing deeper and more detailed insights on the dynamics of some of the most fundamental and relevant optoelectronic processes. There are still many unknown and unexplored areas here. Theory has failed to offer definitive answers. Simulations based on the exact time-dependent Schrödinger equation have not been possible in most cases. Therefore one uses approximations and simpler models to capture the essential physics. Such semi-classical models potentially will help to understand attosecond energy and charge transport in larger molecular systems. Indeed the attoclock provides a unique tool to explore different semi-classical models. For example, the question of whether electron tunneling through an energetically forbidden region takes a finite time or is instantaneous has been subject to ongoing debate for the last sixty years. The tunnelling process, charge transfer, and energy transport all play key roles in electronics, energy conversion, chemical and biological reactions, and fundamental processes important for improved information, health, and energy technologies. We believe the attoclock can help refine and resolve key models for many of these important underlying attosecond processes.

2 319 796 €

Start date: 2013-03-01, End date: 2018-02-28